Paleontology and Geology of Laetoli: Human Evolution in Context
Vertebrate Paleobiology and Paleoanthropology Series Edited by Eric Delson Vertebrate Paleontology, American Museum of Natural History, New York, NY 10024, USA
[email protected] Eric J. Sargis Anthropology, Yale University New Haven, CT 06520, USA
[email protected] Focal topics for volumes in the series will include systematic paleontology of all vertebrates (from agnathans to humans), phylogeny reconstruction, functional morphology, Paleolithic archaeology, taphonomy, geochronology, historical biogeography, and biostratigraphy. Other fields (e.g., paleoclimatology, paleoecology, ancient DNA, total organismal community structure) may be considered if the volume theme emphasizes paleobiology (or archaeology). Fields such as modeling of physical processes, genetic methodology, nonvertebrates or neontology are out of our scope. Volumes in the series may either be monographic treatments (including unpublished but fully revised dissertations) or edited collections, especially those focusing on problem-oriented issues, with multidisciplinary coverage where possible. Editorial Advisory Board Nicholas Conard (University of Tübingen), John G. Fleagle (Stony Brook University), Jean-Jacques Hublin (Max Planck Institute for Evolutionary Anthropology), Ross D.E. MacPhee (American Museum of Natural History), Peter Makovicky (The Field Museum), Sally McBrearty (University of Connecticut), Jin Meng (American Museum of Natural History), Tom Plummer (Queens College/CUNY), Mary Silcox (University of Toronto).
For other titles published in this series, go to www.springer.com/series/6978
Paleontology and Geology of Laetoli: Human Evolution in Context Volume 2: Fossil Hominins and the Associated Fauna
Edited by
Terry Harrison Center for the Study of Human Origins, Department of Anthropology, New York University, 25 Waverly Place, New York, NY 10003, USA
Editor Terry Harrison Center for the Study of Human Origins Department of Anthropology New York University 25 Waverly Place New York, NY 10003 USA
[email protected] ISBN 978-90-481-9961-7 e-ISBN 978-90-481-9962-4 DOI 10.1007/978-90-481-9962-4 Springer Dordrecht Heidelberg London New York © Springer Science+Business Media B.V. 2011 No part of this work may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, microfilming, recording or otherwise, without written permission from the Publisher, with the exception of any material supplied specifically for the purpose of being entered and executed on a computer system, for exclusive use by the purchaser of the work. Cover illustration: Photograph of the L.H. 4 (lectotype) mandible of Australopithecus afarensis superimposed on a view of Laetoli Locality 10 (© and courtesy of Terry Harrison). Printed on acid-free paper Springer is part of Springer Science+Business Media (www.springer.com)
To Australopithecus afarensis for being there when it mattered
wwwwwwwwwwwwwww
Preface
Laetoli in northern Tanzania is one of the most important paleontological and paleoanthropological sites in Africa. It is renowned for the recovery of early hominin fossils belonging to A. afarensis and for the discovery of remarkably well-preserved trails of footprints of hominins. Given the significance of Laetoli for understanding and interpreting the evolutionary history of early hominins the author initiated long-term geological and paleontological investigations at Laetoli and at other fossil localities on the Eyasi Plateau. The overall objectives of the project were to recover additional fossil hominin specimens and to obtain more detailed contextual information on the paleontology, geology, dating, and paleoecology. The field campaigns (1998–2005) have produced important original data on the fossil hominins, their associated fauna, and the paleoecological and paleoenvironmental context. The work presented here is the culmination of that research. It represents the combined effort of a dedicated and experienced field crew who were responsible for collecting the fossils and samples described and analyzed here, and subsequent research by a multidisciplinary team of international specialists. The present volume focuses on the morphology, systematics and paleobiology of the fossil hominins and the associated invertebrate and vertebrate fauna. The companion volume provides an interdisciplinary perspective on the geology, geochronology, paleoecology, taphonomy, paleobotany, and modern-day Serengeti ecosystem. Together, these two volumes present a comprehensive account of the geology, paleontology and paleoecology of Laetoli. It is hoped that the research presented here will provide an important building block in a broader understanding of early hominin evolution, faunal diversity and ecological change in East Africa during the Pliocene, and provide the basis for analyzing early hominin adaptation within the context of broader macroevolutionary models of speciation, diversification and extinction. A special thanks goes to all of the dedicated team members who participated in the expeditions to Laetoli that contributed to the recovery of the material discussed and analyzed here (they are identified individually in the introductory chapter in Volume 1). I am especially grateful to the graduate students (current and former) who participated in the project, often under difficult conditions, and I fully acknowledge their significant contributions to the success of the project. The students who accompanied me into the field were as follows: E. Baker, S. Cooke, C. Fellmann, K. Kovarovic, A. Malyango, L. McHenry, K. McNulty, G. Mollel, C.P. Msuya, T. Rein, C. Robinson, L. Rossouw, M. Seselj, D. Su, M. Tallman and S. Worthington. Of my former graduate students, Denise Su deserves special recognition for taking on the primary role of curating and cataloguing the Leakey and Harrison Laetoli collections at the National Museum of Tanzania in Dar es Salaam, as well as for her valuable assistance with logistics at Laetoli and in Dar es Salaam. I thank the Tanzania Commission for Science and Technology and the Unit of Antiquities in Dar es Salaam for permission to conduct research in Tanzania. Special thanks go to the late Norbert Kayombo (Director General), Paul Msemwa (Director), Amandus Kweka and all of the curators and staff at the National Museum of Tanzania in Dar es Salaam for their support and assistance. I thank the regional, district and ward officers in Arusha Region for their support and hospitality. I am grateful to the Ngorongoro Conservation Area Authority for permission vii
viii
Preface
to conduct research in the conservation area. Emin Korcelik and Naphisa Jahazi of Hertz International in Dar es Salaam arranged the field transportation, and H. Meghji and A. Esmail helped with logistical support in Dar es Salaam. Research at Laetoli benefited from the advice, discussion, help and support from numerous individuals, especially the following: P. Andrews, R. Blumenschine, E. Delson, A. Deino, P. Ditchfield, C. Feibel, S. Frost, C. Harrison, T.S. Harrison, D. M. K. Kamamba, O. Kileo, J. Kingston, A. Kweka, J. LeClair, M. G. Leakey, S. Mataro, G. Ole Moita, E. Mbua, L. McHenry, C. P. Msuya, C. S. Msuya, G. Mollel, M. Muungu, O. Mwebi, J. Pareso, C. Peters, M. Pickford, K. Reed, C. Saanane, W. Sanders, C. Swisher, and S. Waane. Bill Sanders deserves special mention for applying his exceptional talents to preparing and casting some of the Laetoli specimens, as does Jen LeClair for her tireless efforts in helping to organize the collections and entering data in the catalogue. I thank the curators and staff at the various museums and repositories for allowing me access to archival materials, fossils and comparative specimens in their care. These include: National Museums of Tanzania, Kenya National Museum, American Museum of Natural History, Natural History Museum in London, Humboldt-Universität Museum für Naturkunde in Berlin, Eberhard-Karls Universitat Tübingen Institut für Ur- und Frühgeschichte und Archäologie des Mittelalters and Institut und Museum für Geologie and Paläontologie. The following individuals provided critical comments and advice about the research presented in this volume and its companion: A. Alexandre, P. Andrews, M. Anton, M. Avery, M. Bamford, F. Bibi, L. Bishop, R, Bobe, R. Bonnefille, F, Brown, P. Butler, C. Crumly, A. Deino, P. Ditchfield, P. Duringer, M. Erbajeva, R. Evander, C. Feibel, Y. Fernandez-Jalvo, B. Fine-Jacobs, L. Flynn, S. Frost, T, Furman, J. Genise, A. Gentry, D. Geraads, H. Gilbert, U. Goehlich, J.H. Harris, K. Heissig, A. Hill, P. Holroyd, D. Iwan, N. Jablonski, J. Kappelman, T. Kaiser, R. Kay, J. Kingdon, J. Kingston, W. Kimbel, J. Knott, K. Kovarovic, N. Kristensen, O. Kullmer, F. de Lapparent de Broin, M. Lewis, N. Lopez-Martinez, S. Manchester, I. MacDougall, L. McHenry, S. McNaughton, K. Metzger, P. Meylan, C. Mourer-Chauviré, R. Oberprieler, E. O’Brien, D. Parmley, M. Pavia, C. Peters, M. Pickford, I. Poole, B. Ratcliffe, D. Reed, K. Reed, W.J. Sanders, M. Sponheimer, D. Su, Z. Szyndlar, R. Tabuce, P. Tassy, B. Tiffney, J. van der Made, A. Vincens, C. Ward, H. Wesselman, E. Wheeler, and A. Winkler. Special thanks go to Terri Harrison, Chris Harrison and Leahanne Sarlo for their assistance with many aspects of the editorial process. I thank Eric Delson, Eric Sargis and the Editorial and Production team at Springer, especially Tamara Welschot and Judith Terpos. Fieldwork at Laetoli and subsequent research was supported by grants from the National Geographic Society, the Leakey Foundation, and the National Science Foundation (Grants BCS-0216683 and BCS-0309513). New York
Terry Harrison
Contents
1 Introduction: The Laetoli Hominins and Associated Fauna................................ Terry Harrison
1
2 Rodents...................................................................................................................... Christiane Denys
15
3 The Lower Third Premolar of Serengetilagus praecapensis (Mammalia: Lagomorpha: Leporidae) from Laetoli, Tanzania.......................... Alisa J. Winkler and Yukimitsu Tomida
55
4 Macroscelidea........................................................................................................... Alisa J. Winkler
67
5 Galagidae (Lorisoidea, Primates)........................................................................... Terry Harrison
75
6 Cercopithecids (Cercopithecidae, Primates)......................................................... Terry Harrison
83
7 Hominins from the Upper Laetolil and Upper Ndolanya Beds, Laetoli........................................................................................................................ 141 Terry Harrison 8 Carnivora.................................................................................................................. 189 Lars Werdelin and Reihaneh Dehghani 9 Proboscidea............................................................................................................... 233 William J. Sanders 10 Orycteropodidae...................................................................................................... 263 Terry Harrison 11 Rhinocerotidae......................................................................................................... 275 Elina Hernesniemi, Ioannis X. Giaourtsakis, Alistair R. Evans, and Mikael Fortelius 12 Equidae..................................................................................................................... 295 Miranda Armour-Chelu and Raymond L. Bernor 13 Suidae........................................................................................................................ 327 Laura C. Bishop
ix
x
14 Giraffidae.................................................................................................................. 339 Chris A. Robinson 15 Bovidae...................................................................................................................... 363 Alan W. Gentry 16 Amphibia and Squamata......................................................................................... 467 Jean-Claude Rage and Salvador Bailon 17 Tortoises (Chelonii, Testudinidae).......................................................................... 479 Terry Harrison 18 Aves............................................................................................................................ 505 Antoine Louchart 19 Beetles (Insecta: Coleoptera).................................................................................. 535 Frank-T. Krell and Wolfgang Schawaller 20 Lepidoptera, Insecta................................................................................................ 549 Ian J. Kitching and S. Sadler 21 Trace Fossils Interpreted in Relation to the Extant Termite Fauna at Laetoli, Tanzania.................................................................................................. 555 Johanna P.E.C. Darlington 22 Gastropoda............................................................................................................... 567 Peter Tattersfield Index.................................................................................................................................. 589
Contents
Contributors
Miranda Armour-Chelu Department of Anatomy, Laboratory of Evolutionary Biology, College of Medicine, Howard University, 520 W St, N.W. Washington, DC 20059, USA
[email protected] Salvador Bailon CNRS UMR 7194-7209, MNHN, CP 55, 55 Rue Buffon, 75005, Paris, France
[email protected] Raymond L. Bernor Department of Anatomy, Laboratory of Evolutionary Biology, College of Medicine, Howard University, 520 W St, N.W. Washington, DC 20059, USA
[email protected] Laura C. Bishop Research Centre in Evolutionary Anthropology and Palaeoecology, School of Natural Sciences and Psychology, Liverpool John Moores University, Byrom St, Liverpool L3 3AF, UK
[email protected] Johanna P.E.C. Darlington University Museum of Zoology, Downing Street, Cambridge, CB2 3EJ, UK
[email protected] Reihaneh Dehghani Department of Palaeozoology, Swedish Museum of Natural History, Box 50007 S-104 05, Stockholm, Sweden
[email protected] Christiane Denys Department of Systematics and Evolution – CP51, UMR7205 CNRS: Origine structure & évolution de la Biodiversité, MNHN, 55 rue Buffon, 75005, Paris, France
[email protected] Alistair R. Evans School of Biological Sciences, Monash University, Victoria 3800, Australia
[email protected] Mikael Fortelius Department of Geosciences and Geography, University of Helsinki, P.O. Box 64 FIN-00014, Helsinki, Finland
[email protected] xi
xii
Alan W. Gentry c/o Department of Palaeontology, Natural History Museum, Cromwell Road, London, SW7 5BD, UK
[email protected] Ioannis X. Giaourtsakis Department of Geo- and Environmental Sciences, Section of Paleontology, Ludwig-Maximilians-University of Munich, Richard-Wagner-Strasse 10, D-80333 Munich, Germany
[email protected] Terry Harrison Center for the Study of Human Origins, Department of Anthropology, New York University, 25 Waverly Place, New York, NY 10003, USA
[email protected] Elina Hernesniemi Department of Geosciences and Geography, University of Helsinki, P.O. Box 64 FIN-00014, Helsinki, Finland
[email protected] Ian J. Kitching Department of Entomology, The Natural History Museum, Cromwell Road, London, SW7 5BD, UK
[email protected] Frank-T. Krell Denver Museum of Nature & Science, 2001 Colorado Blvd. Denver, CO 80205, USA
[email protected] Antoine Louchart Institut de Génomique Fonctionnelle de Lyon, team “Evo-devo of vertebrate dentition”, Université de Lyon, Université Lyon 1, CNRS UMR 5242, INRA, Ecole Normale Supérieure de Lyon, 46 Allée d’Italie, 69364 Lyon cedex 07, France
[email protected];
[email protected] Jean-Claude Rage Département Histoire de la Terre, CNRS UMR 7207, MNHN, CP 38, 8 rue Buffon, 75231, Paris cedex 05, France
[email protected] Chris A. Robinson Department of Biology, Bronx Community College, 2155 University Avenue, Bronx, NY 10453, USA
[email protected] S. Sadler Department of Entomology, The Natural History Museum, Cromwell Road, London, SW7 5BD, UK William J. Sanders Museum of Paleontology, The University of Michigan, 1109 Geddes Avenue, Ann Arbor, MI 48109, USA
[email protected] Wolfgang Schawaller Staatliches Museum für Naturkunde, Rosenstein 1, D-70191, Stuttgart, Germany
[email protected] Contributors
Contributors
xiii
Peter Tattersfield Department of Biodiversity and Systematic Biology, National Museum of Wales, Cathays Park, Cardiff, CF10 3NP, UK
[email protected] Yukimitsu Tomida National Museum of Nature and Science, 3-23-1 Hyakunincho, Shinjukuku, Tokyo 169-0073, Japan
[email protected] Lars Werdelin Department of Palaeozoology, Swedish Museum of Natural History, Box 50007 S-104 05 Stockholm, Sweden
[email protected] Alisa J. Winkler Roy M. Huffington Department of Earth Sciences, Southern Methodist University, Dallas, TX 75275, USA; Department of Cell Biology, University of Texas Southwestern Medical Center at Dallas, Dallas, TX 75390, USA
[email protected] wwwwwwwwwwwwwwwwwwww
Chapter 1
Introduction: The Laetoli Hominins and Associated Fauna Terry Harrison
Abstract Laetoli in northern Tanzania is one of the most important paleontological and paleoanthropological localities in Africa. In addition to fossil hominins, there is a diverse associated fauna. The Laetoli fauna is important because it serves as a key comparative reference for other PlioPleistocene sites in Africa, it samples several time periods that are generally poorly represented at other paleontological sites in East Africa, and it provides key insights into the faunal and floral diversity during the Pliocene. As a result of renewed fieldwork at Laetoli (1998–2005) more than 25,000 fossils have been collected, of which more than half are fossil mammals. Most of the fossils were recovered from the Upper Laetolil Beds (3.6–3.85 Ma), but smaller samples came from the Lower Laetolil Beds (3.85–4.4 Ma) and Upper Ndolanya Beds (2.66 Ma). These include new specimens of Australopithecus afarensis from the Upper Laetolil Beds and the first finds of fossil hominins from the Upper Ndolanya Beds, attributable to Paranthropus aethiopicus. Inferences about the paleoecology at Laetoli are important for understanding the possible range of hominin habitat preferences and ecological change in East Africa during the Pliocene. The evidence from a wide range of analyses indicates that a mosaic of closed woodland, open woodland, shrubland and grassland dominated the paleoecology of the Upper Laetolil Beds. The region would have been dry for most of the year, except for the possible occurrence of permanent springs along the margin of the Eyasi Plateau and ephemeral pools and rivers during the rainy season. The paleoecological reconstruction of the Upper Ndolanya Beds is more problematic because of conflicting lines of evidence, but it is very likely that conditions were drier than in the Upper Laetolil Beds with a greater proportion of grassland, but that closed and open woodlands were still a major part of the ecosystem. Keywords Pliocene • Laetolil Beds • Ndolanya Beds • Fauna • Paleontology • Paleoecology
T. Harrison (*) Center for the Study of Human Origins Department of Anthropology, New York University, 25 Waverly Place, New York, NY 10003, USA e-mail:
[email protected] Introduction Laetoli in northern Tanzania is one of the most important paleontological and paleoanthropological localities in Africa. The site is well known for the recovery of fossils of the early hominin Australopithecus afarensis, as well as trails of hominin footprints. The associated fauna from Laetoli is very diverse (Leakey and Harris 1987), with over 100 species of mammals identified, along with the remains or traces of fossil amphibians, reptiles, birds, insects, gastropods and plants. As such, it serves as a key reference fauna, one that is reliably dated, for comparisons with other Plio-Pleistocene sites in Africa. Equally importantly, the Upper Laetolil Beds (3.6–3.85 Ma) and Upper Ndolanya Beds (2.66 Ma) sample time periods that are generally poorly represented at other paleontological sites in East Africa, and the fossils from these stratigraphic units provides key insights into the faunal and floral diversity during the Pliocene. Detailed information on the paleontological localities and geology at Laetoli is presented in the companion volume (Harrison 2011a), but the essential information is summarized in Figs. 1.1–1.5. Laetoli is unusual among sites in East Africa in the absence of sedimentological or paleontological evidence for extensive and/or permanent bodies of water, and in having an inferred paleoecological setting that is less extensively wooded than its penecontemporaneous sites. Given these distinctive characteristics of the paleoecology at Laetoli, the site provide an important building block for inferring the possible range of hominin habitat preferences and for understanding ecological change in East Africa during the Pliocene and its impact on early human evolution. As a consequence, the ecological context at Laetoli has been extensively investigated in the past (Leakey and Harris 1987; Andrews 1989, 2006; Cerling 1992; Andrews and Humphrey 1999; Musiba 1999; Kovarovic et al. 2002; Kovarovic 2004; Su 2005; Harrison 2005; Kovarovic and Andrews 2007; Kingston and Harrison 2007; Musiba et al. 2007; Su and Harrison 2007, 2008; Andrews and Bamford 2008; Peters et al. 2008), and is a special focus of renewed investigations since 1998.
T. Harrison (ed.), Paleontology and Geology of Laetoli: Human Evolution in Context. Volume 2: Fossil Hominins and the Associated Fauna, Vertebrate Paleobiology and Paleoanthropology, DOI 10.1007/978-90-481-9962-4_1, © Springer Science+Business Media B.V. 2011
1
2
T. Harrison
Fig. 1.1 A sketch map of the Eyasi Plateau showing the major rivers and villages, as well as the three main paleontological research areas: Laetoli, Kakesio and Esere-Noiti (see Figs. 1.2–1.4 for detail of insets) (From Harrison and Kweka 2011)
The Laetoli Fauna During the course of renewed fieldwork at Laetoli, between 1998 and 2005, more than 25,000 fossils have been collected (Table 1.1). These consist mainly of fossil mammals (58.1%), but also include the remains of birds (4.9%), reptiles and amphibians (1.9%), invertebrates (33.3%) and plants (1.8%) (Table 1.2). Most of the fossil mammals were recovered from the Upper Laetolil Beds, but smaller samples came from the Lower Laetolil Beds and Upper Ndolanya Beds. Represen tative fossil vertebrates were also recovered from the Olpiro and Ngaloba Beds, but no systematic collections were made from these stratigraphic units. Renewed investigations at Laetoli have led to the recovery of additional fossil hominins (Harrison 2011b). These include further specimens attributable to A. afarensis from the Upper Laetolil Beds, and provide the basis, along with other previously undescribed specimens, for a reassessment of the morphology and evolutionary status of the A. afarensis sample from Laetoli. In addition, two hominins were recovered from the Upper Ndolanya Beds, and these represent the first homi-
nins from this stratigraphic unit. A maxilla from the Upper Ndolanya Beds at Silal Artum (EP 1500/01) is important because it represents the only specimen of Paranthropus aethiopicus recovered from outside the Turkana basin, and it is among the oldest securely dated specimens definitively attributable to this taxon (Harrison 2011b). The contributions in Leakey and Harris (1987) provided the last comprehensive account of the systematics of the Laetoli fauna. Since that time, however, there have been major advances in our understanding of the systematics and paleobiology of late Miocene and Plio-Pleistocene faunas of Africa, as well as many reports of new localities and faunas. Renewed investigations at Laetoli have allowed a thorough revision of the systematics of the Laetoli fauna, along with a greater emphasis on understanding the paleobiology of the fauna and its paleoecological implications. All of the mammalian taxa have been restudied, with the exception of the Camelidae and Chalicotheriidae (their analysis is still ongoing). The faunal list now includes nine new species of mammals and six new species of invertebrates, all of which are described in this volume. In addition, one new species of
1 Hominins and Associated Fauna
3
Fig. 1.2 Map of the Laetoli area showing the main outcrops of the Upper Laetolil and Upper Ndolanya Beds and the paleontological collecting localities (From Harrison and Kweka 2011)
ostrich, Struthio kakesiensis, has been named previously (Harrison and Msuya 2005), based on new collections from the Lower and Upper Laetolil Beds. Mary Leakey’s team did recover a small sample of fossil vertebrates from the Lower Laetolil Beds at Kakesio early in their campaign, but the most intensive phase of research at the site took place in 1982, and as a result most of the fossil material and geological information obtained were not included in the Laetoli monograph (Leakey and Harris 1987). Harris (1987) published a brief summary of the fauna from the Lower Laetolil Beds, but most of the specimens remained undescribed. The specimens have been incorporated in the current studies of the fauna. The new collection of fossil mammals from the Lower Laetolil Beds is small (n = 251 specimens), but with more intensive prospecting, especially in the areas of Kakesio and Noiti, it would
be possible to recover a much larger sample. Given the age of the Lower Laetolil Beds (3.85–4.4 Ma), the fauna from these beds could be extremely important in the study of human evolution, because it samples the time period between the last occurrence of Ardipithecus and the first appearance of Australopithecus. The Lower Laetolil fauna now includes 27 species of mammals (up from 18 in 1987) (Table 1.3). It is dominated by bovids, equids and proboscideans. Small mammals are rare, and there is a strong taphonomic bias in favor of large mammals. Most of the mammalian taxa (78%) in the Lower Laetolil Beds also occur in the Upper Laetolil Beds, implying a strong biogeographic provinciality, despite the time difference (Table 1.3). However, several species occur in the Lower Laetolil Beds that are not present in the Upper Laetolil Beds. These include: Anancus kenyensis, Petromus sp., Heterocephalus manthii, aff.
4
T. Harrison
Fig. 1.3 Map of the Kakesio area showing the main outcrops of the Lower Laetolil Beds and the paleontological collecting localities (grey shaded areas) (From Harrison and Kweka 2011)
Fig. 1.4 Map of the Esere-Noiti area showing the main outcrops of the Lower Laetolil Beds and the paleontological collecting localities (grey shaded areas) (From Harrison and Kweka 2011)
Proteles, Aonyxini gen. et sp. nov., and possibly Gazella granti (Sanders 2011; Denys 2011; Werdelin and Dehghani 2011; Gentry 2011). Most of these are very rare taxa (just one or a few specimens), with the exception of Anancus kenyensis. There are now 85 species of mammals recorded from the Upper Laetolil Beds (compared with 71 in 1987) (Table 1.3). Including the Harrison and Leakey collections combined there are now over 18,000 mammal specimens known from the Upper Laetolil Beds (Table 1.3). The large mammal fauna is dominated by bovids (34% of all mammal specimens), with Madoqua avifluminis, Parmularius pandatus and Gazella janenschi being especially common (Gentry 2011). At most East African localities Neotragini are rare, whereas at Laetoli Madoqua is the by far the most common bovid taxon. Giraffids, with three species of different sizes represented, are also quite common (6.3% of all mammal specimens). Micromammals are well-represented in the Upper Laetolil Beds, especially the lagomorph Serengetilagus praecapensis, which is the commonest species, occurring ubiquitously throughout the unit (Denys 2011; Winkler and Tomida 2011). However, there is a high likelihood that small species of
1 Hominins and Associated Fauna
5 Table 1.2 Number of specimens and the frequency of fossil mammals collected at Laetoli and other localities on the Eyasi Plateau from 1998 to 2005 Lower Laetolil Upper Laetolil Upper Ndolanya Beds Beds Beds
Fig. 1.5 Simplified stratigraphic scheme of Laetoli sediments showing the main stratigraphic units (left) and the chronology (right, Ma = megaannum) (After Hay 1987; Ndessokia 1990; Manega 1993; Ditchfield and Harrison 2011; Deino 2011)
Table 1.1 Number of fossils collected 1998–2005 Taxon LLB ULB UNB Total
% of total
Mammals 258 12,383 2,378 15,019 58.1 Birdsb 3 185 9 197 0.8 Struthioc 427 343 289 1,059 4.1 Reptiles and 103 352 34 489 1.9 amphibiansd 290 4,612 282 5,184 20.1 Molluskse Insectsf 460 1,857 1,103 3,420 13.2 Plantsg 7 457 4 468 1.8 Total 1,548 20,189 4,095 25,832 100.0 Specimen counts do not include fossils from the Olpiro or Ngaloba Beds LLB Lower Laetolil Beds, ULB Upper Laetolil Beds, UNB Upper Ndolanya Beds a
For more detailed information on fossil mammals see Table 1.2 Includes bones and eggs, except for those assigned to Struthio c Egg shell fragments of Struthio d Mostly the remains of tortoises, but the count does include snakes, lizards and amphibians e Terrestrial gastropods (For more detailed data on specimen counts see Tattersfield 2011) f Mainly consists of cocoons and brood cells of solitary bees, but also includes casts of insects, termitaries, and brood cells of dung beetles g Includes wood, twigs, leaves, and seeds (see Bamford 2011a, b) a
b
Taxon
N
%
N
Macroscelididae Galagidae Cercopithecidae Hominidae Rodentia Leporidae Carnivora Proboscidea Orycteropodidae Equidae Rhinocerotidae Chalicotheriidae Suidae Camelidae Giraffidae Bovidae Total
0 0 1 0 10 15 13 37 1 55 21 0 12 0 8 79 252
0 0 0.40 0 3.97 5.95 5.16 14.68 0.40 21.83 8.33 0 4.76 0 3.17 31.35 100.0
4 0.03 0 0 1 0.01 0 0 111 0.91 1 0.04 2 0.02 2 0.09 855 7.00 104 4.55 4,640 38.00 398 17.41 424 3.47 54 2.36 158 1.29 24 1.05 26 0.21 2 0.09 330 2.70 110 4.81 473 3.87 29 1.27 3 0.02 0 0 244 2.00 27 1.18 26 0.21 6 0.26 772 6.32 70 3.06 4,145 33.95 1,459 63.82 12,214 100.01 2,286 99.99
%
N
%
rodents are under-represented in the collections due to taphonomic and collecting biases (Denys 2011; Reed and Denys 2011). Primates, including hominins, are rare, and comprise less than 1% of the mammalian fauna (Harrison 2011b, c, d; Table 1.2). Fossil mammals are also abundant in the Upper Ndolanya Beds, which are separated in time from the Upper Laetolil Beds by a hiatus of about one million years. Forty-nine species of mammals are currently recognized (up from 38 species in 1987) (Table 1.3). Of these, just over half of the species (53%) are shared with the Upper Laetolil Beds. However, there is apparently a significant faunal turnover between these two units (between 3.6 and 2.66 Ma). Among the large mammals, Eurygnathohippus aff cornelianus replaces Eurygnathohippus aff. hasumense, and Ceratotherium simum, Metridiochoerus andrewsi, Giraffa pygmaea, and Camelus sp. make their first appearance in the Laetoli fauna (Armour-Chelu and Bernor 2011; Hernesniemi et al. 2011; Bishop 2011; Robinson 2011). Several bovids also appear for the first time, including Parmularius altidens, Parmularius parvicornis, Megalotragus sp., Tragelaphus sp. cf. T. buxtoni and Antidorcus recki (Gentry 2011). Among the micromammals Gerbilliscus winkleri replaces G. satimani, and Thryonomys wesselmani appears for the first time (Denys 2011). There is also an important difference in the hominins, with Australopithecus afarensis being replaced by Paranthropus aethiopicus (Harrison 2011b). A better understanding of the ecological differences and changes in the community structure between the Upper Laetolil Beds and the Upper Ndolanya Beds should provide important clues to
6
T. Harrison Table 1.3 List of the fauna from the main stratigraphic units at Laetoli Class Order Family Genus and species Insecta
Hymenoptera Coleoptera
Indeterminate Tenebrionidae
Scarabaeidae
Gastropoda
Diptera Lepidoptera Isoptera
Indeterminate Saturniidae Termitidae
Pulmonata
Indeterminate Succineidae Cerastidae
Subulinidae
Vertiginidae Streptaxidae
Achatinidae
Urocyclidae
Bradybaenidae
LLB
Tentyriini sp. A (?Tentyria) Tentyriini sp. B ?Tentyriini sp. C Molurini sp. A (?Arturium) Calcitoryctes magnificus Melolonthinae: Schizonychini, sp. A Coprinisphaera ndolanyanus Coprinisphaera laetoliensis Lazaichnus amplus
UNB X
X X X X X
X X X X X X X
Bunaeini indet. Macrotermes spp. Apicotermitinae indet. “Succinea” sp. A Gittenedouardia laetoliensis Cerastus sp. A Subulona pseudinvoluta Pseudoglessula (Kempioconcha) aff. gibbonsi Kenyaella leakeyi Kenyaella harrisoni Subuliniscus sp. A Pupoides coenopictus Streptostele (Raffraya) aff. horei Streptostele sp. A Gulella sp. A Burtoa nilotica Limicolaria martensiana Achatina (Lissachatina) indet. Trochonanina sp. B Urocyclinae sp. A Urocyclinae sp. B Urocyclinae sp. C Urocyclinae sp. D Urocyclinae sp. E Urocyclinae sp. F Halolimnohelix rowsoni
ULB X X
X X X
X
X X X
?
X X X
X X
X X
X X X X
X
X
X
X X X X X X X X
X
X X X X X X X
(continued)
1 Hominins and Associated Fauna
7
Table 1.3 (continued) Class Order
Family
Amphibia Reptila
Indeterminate Testudinidae
Anura Chelonii
Crocodilia Squamata
Crocodylidae Acrodonta indet. Scincomorpha indet. Boidae Colubridae
Elapidae Viperidae
Aves
Struthioniformes Struthionidae Galliformes
Phasianidae
Numididae
Ciconiiformes
Ardeidae
Charadriiformes Scolopacidae Accipitriformes Accipitridae
Falconiformes Columbiformes
Strigiformes
Coliiformes Passeriformes
Falconidae Columbidae
Tytonidae Strigidae
Collidae Indeterminate
Genus and species
LLB
ULB
UNB
Stigmochelys brachygularis “Geochelone” laetoliensis Crocodylus sp.
X
X X
X
X
X
X X X
Python sebae or P. natalensis cf. Thelotornis sp. cf. Rhamphiophis sp. Indeterminate sp. A Indeterminate sp. B Naja robusta ?indeterminate sp. X Bitis sp. nov. or Bitis olduvaiensis Struthio kakesiensis X Struthio camelus Francolinus sp. A aff. F. (Peliperdix) sephaena Francolinus (Pternistis) sp. B cf. Francolinus sp. indet. cf. Agelastes sp. Numida/Guttera sp. X Acryllium vulturinum cf. Ardea sp. Aegypius sp. Calidrinae indet. cf. Buteo sp. Aquilini indet. sp. A cf. Aquilini indet sp. B Falco cf. eleonorae Falconiformes indet. Columba sp. (sp. A) Streptopelia sp. (sp. B) Columbidae indet. (sp. C) Tyto sp. Bubo cf. lacteus (sp. A) Asio sp. (sp. B) cf. Strigidae (sp. C) Colius sp. cf. Passerida indet.
X X X X X X X X
X
X X X
X X
X
X
X X X X
X
X X X X X X X X X X
X X
X X X X X X (continued)
8
T. Harrison Table 1.3 (continued) Class Mammalia
Order Macroscelidea
Family Macroscelididae
Tubulidentata Proboscidea
Orycteropodidae Deinotheriidae Gomphotheriidae Stegodontidae
Elephantidae
Primates
Galagidae Cercopithecidae
Hominidae
Rodentia
Sciuridae
Cricetidae
Muridae
Thryonomyidae Petromuridae Bathyergidae
Hystricidae
Pedetidae Lagomorpha
Leporidae
Soricimorpha
Soricidae
Genus and species Rhynchocyon pliocaenicus Orycteropus sp. Deinotherium bozasi Anancus kenyensis Anancus ultimus Stegodon sp. cf. Stegodon kaisensis Loxodonta sp. cf. Loxodonta cookei Loxodonta exoptata Laetolia sadimanensis Parapapio ado Papionini indet. cf. Rhinocolobus sp. Cercopithecoides sp. Australopithecus afarensis Paranthropus aethiopicus Paraxerus meini Xerus sp. Xerus janenschi Gerbilliscus satimani Gerbilliscus winkleri Gerbilliscus cf. inclusus Dendromus sp. Steatomys sp. Saccostomus major Saccostomus sp. Aethomys sp. Thallomys laetolilensis Mastomys cinereus Mus sp. Thryonomys wesselmani Petromus sp. Heterocephalus quenstedti Heterocephalus manthii Hystrix leakeyi Hystrix makapanensis Xenohystrix crassidens Pedetes laetoliensis Pedetes sp. Serengetilagus praecapensis ?Crocidura sp.
LLB
ULB X
X
X X
X X
UNB
?
X X
X
X
X X
X
X X X X X
X X
X X X X X
X X
X X
cf.
X X X X X
cf. X X
X X X X X X X X
X
X X X
X
X X X (continued)
1 Hominins and Associated Fauna
9
Table 1.3 (continued) Class
Order
Family
Genus and species
Carnivora
Canidae
?Nyctereutes barryi cf. Canis sp. A cf. Canis sp. B aff. Otocyon sp. Prepoecilogale bolti Mellivora sp. Aonyxini gen. et sp. nov. Mustelidae indet. Viverra leakeyi Genetta sp. aff. Viverridae Herpestes palaeoserengetensis Herpestes ichneumon Galerella sp. Helogale palaeogracilis Mungos dietrichi Mungos sp. nov. Crocuta dietrichi Parahyaena howelli Ikelohyaena cf. I. abronia Lycyaenops cf. L. silberbergi ?Pachycrocuta sp. aff. Proteles sp. Dinofelis petteri Homotherium sp. Panthera sp. aff. P. leo Panthera sp. cf. P. pardus Acinonyx sp. Caracal sp. or Leptailurus sp. Felis sp. Eurygnathohippus aff. hasumense Eurgnathohippus aff. cornelianus Ancylotherium hennigi Ceratotherium efficax Ceratotherium cf. simum Ceratotherium sp. Diceros sp. Notochoerus euilus Notochoerus jaegeri Nyanzachoerus kanamensis Potamochoerus afarensis
Mustelidae
Viverridae
Herpestidae
Hyaenidae
Felidae
Perissodactyla
Equidae
Chalicotheriidae Rhinocerotidae
Artiodactyla
Suidae
LLB
ULB X X X X X X
UNB
X
X X X X X X
X
X
X
X X X X X X X
X X X ?
X X X
X
X X X
X X
X
X
X X
X
X X
X
X X X
X
X X X
X X
X X X X X (continued)
10
T. Harrison Table 1.3 (continued) Class
Order
Family
Giraffidae
Camelidae Bovidae
understanding the differentiation of the Paranthropus lineage. The mammalian fauna from the Upper Ndolanya Beds is heavily skewed towards bovids (63.8% of all specimens), especially medium- and large-sized alcelaphines, probably as a consequence of an unusual combination of taphonomic
Genus and species Kolpochoerus heseloni Metridiochoerus andrewsi Giraffa stillei Giraffa jumae Giraffa pygmaea Sivatherium maurusium Camelus sp. Tragelaphus sp. Tragelaphus sp. cf. T. buxtoni Simatherium kohllarseni Brabovus nanincisus Bovini sp. indet. Cephalophini sp. Hippotragus sp. Hippotragus sp. aff. cookei? Oryx deturi Oryx sp. Parmularius pandatus Parmularius ?altidens Parmularius parvicornis Alcelaphini, larger sp. indet. Alcelaphini, small sp. Megalotragus kattwinkeli or M. isaaci ?Connochaetes sp. Reduncini sp. indet. Madoqua avifluminis ?Raphicerus sp. Aepyceros dietrichi Aepyceros sp. “Gazella” kohllarseni Gazella janenschi Gazella granti Gazella sp. Antidorcas recki
LLB
ULB
UNB
X
X X
aff.
X aff.
X
X
aff. aff. aff. X
X X X X X X
X X X
X ? X
X X X
X X X
X
X ? X
X
X X
X X
X X
X X X X X
X X X
X ?
X ? X X
factors (Table 1.2). The other common species in the Upper Ndolanya fauna is Serengetilagus praecapensis (17.4% of all mammal specimens) (Table 1.2). In addition to fossil mammals, study of the non-mammalian fauna and paleobotanical remains are essential for a complete
1 Hominins and Associated Fauna
understanding and appreciation of the biotic diversity and paleoecology at Laetoli during the Pliocene. These investigations include the first detailed studies to be undertaken of the fossil insects, gastropods, birds, lizards and snakes from Laetoli (Krell and Schawaller 2011; Kitching and Sadler 2011; Tattersfield 2011; Louchart 2011; Rage and Bailon 2011). Research on the fossil ostriches and birds’ eggs has already been published (Harrison 2005; Harrison and Msuya 2005). The contributions presented in this volume provide the basis for a major systematic revision of the Laetoli fauna, as well as a much better appreciation of the paleobiology and paleoecology. The fossil wood and other paleobotanical remains, which provide important insights into the paleoecology of Laetoli, are described in the companion volume (Bamford 2011a, b; Rossouw and Scott 2011).
Paleoecology A major focus of renewed investigations at Laetoli has been on reconstructing the paleoecology (Harrison 2011e). Study of the paleoecology provides critical evidence for understanding the context of early hominin evolution. It also allows researchers to pose important questions about hominin habitat preferences, ecology and paleobiology, and to include these data in larger-scale macroevolutionary models of speciation, biogeography, diversification and extinction. With these questions in mind, renewed work at Laetoli has attempted to reconstruct the paleoecology using information from a wide diversity of sources (i.e., modern-day ecosystems, paleobotany, phytoliths, palynology, invertebrate and invertebrate paleontology, stable isotopes, mesowear, ecomorphology, and community structure analyses) (Andrews et al. 2011; Bamford 2011a, b; Rossouw and Scott 2011; Kingston 2011; Kaiser 2011; Hernesniemi et al. 2011; Harrison 2011f; Bishop et al. 2011; Kovarovic and Andrews 2011; Su 2011; Reed 2011; Reed and Denys 2011). The geological and sedimentological evidence indicates that the Laetoli area had a relatively low topography during the Pliocene, with a gently undulating terrain. There is evidence of rivers and streams in the Upper Laetolil Beds, probably with a greater extent and capacity than the present-day hydrological system, but these rivers only flowed during the wet season, and were dry for most of the year (Ditchfield and Harrison 2011). The watercourses originated in the volcanic highlands about 20 km to the east, and flowed southwest across the Laetoli area, and it is likely that they drained into the developing Eyasi basin. This network of watercourses would have supported a complex vegetational mosaic, including dense stands of riverine woodland and bushland (Ditchfield and Harrison 2011). Ephemeral ponds and small lakes would have dotted the landscape during the rainy sea-
11
son, but these would have dried up during the dry season. There is no evidence of large permanent bodies of water in the Upper Laetolil Beds or Upper Ndolanya Beds, and this is consistent with the absence of aquatic and hydrophilic vertebrates (i.e., hippopotamids, crocodiles, turtles and fishes), with the exception of rare finds of anurans (Rage and Bailon 2011; Ditchfield and Harrison 2011). The paleoenvironment of the Lower Laetolil Beds appears to have been similar to that of the Upper Laetolil Beds and Upper Ndolanya Beds, but there is better evidence of shallow pools and lakes. Aquatic vertebrates are extremely rare in the Lower Laetolil Beds, but the fauna does include an otter and there is also single confirmed specimen of Crocodylus. Harris (1987) reported the presence of fish from the Lower Laetolil Beds, but this has not been confirmed, and their record has been removed from the revised faunal list (Table 1.3). The very common traces of termite bioturbation, burrows of solitary hymenoptera, and the occurrence of aestivating gastropods throughout the Laetoli sequence, all point to widespread paleosols that were well drained and free from inundation for much of the year. It is very likely that run-off from the volcanic highlands would have continued yearround, with water flowing below the surface even during the dry season, just as it does today. Presently, springs occur along the edge of the Eyasi escarpment where the Laetolil Beds interface with the underlying impervious Precambrian basement rocks, and these provide a permanent source of water for wildlife and the local inhabitants. Given that similar geomorphological and topographic features were in place during the Pliocene, it is likely that springs were present in the Laetoli area, and that these offered an important source of water during the long dry season in what would otherwise have been a relatively dry and waterless terrain. Ash fall deposits periodically blanketed the Laetoli area, forming distinctive marker tuffs in the Upper Laetolil Beds. These heavy inundations of carbonatite volcanic ash would have had an adverse effect on the local ecosystem, including burial of the ground vegetation and making standing bodies of water toxic (Peters et al. 2008). The subsequent formation of calcretes and hardpans would have led to a landscape dominated by grasslands and open woodlands. However, these periods of disruption were apparent relatively short-lived, and the climax vegetation would have quickly re-established itself. The ash falls in the Lower Laetolil Beds were thicker and more frequent than in the Upper Laetolil Beds, and undoubtedly would have caused more dramatic short-term disruptions to the local ecosystem. However, the greater degree of fluvial reworking and bioturbation of the Lower Laetolil Beds indicates that the sediments quickly formed weakly developed paleosols that could have supported rapid re-establishment of the climax vegetation. The paleobotanical evidence provides important clues to reconstructing the paleoecology at Laetoli. The fossil wood
12
from the Lower Laetolil Beds at Noiti suggests that woodlands and forest covered the lower slopes of the volcanic highlands to the east of Laetoli, and that a mosaic of woodland, bushland and wooded grasslands occurred more distally (Bamford 2011a). Plant macrofossils from the Upper Laetolil Beds suggest a diverse flora, with vegetation that included forest and woodland elements (Bamford 2011b). The study of the phytoliths indicates that grasses were common at Laetoli during the Pliocene, but they were probably not the dominant vegetation type (Rossouw and Scott 2011). The Lower Laetolil Beds appear to have been deposited in a relatively mesic habitat dominated by C3 grasses. Conditions became drier during the lower part of the Upper Laetolil Beds and more mesic conditions prevailed again during the upper part, with a shift from C3 dominated grasses to C4 dominated grasses. The phytolith evidence indicates that the paleoecology of Upper Ndolanya Beds was one of relatively arid grasslands, dominated by C4 grasses. Studies of the stable isotopes, mesowear, bovid postcranial ecomorphology, small and large mammal community structure, and the bird fauna provide a picture of the Laetoli paleoecology that is largely consistent with that of the paleobotanical evidence (Kingston 2011; Kaiser 2011; Bishop et al. 2011; Su 2011; Denys 2011; Hernesniemi et al. 2011; Louchart 2011; Kovarovic and Andrews 2011). The ecology during deposition of the Upper Laetolil Beds was a vegetational mosaic with woodland, bushland and grasslandsavanna. The ungulate fauna was dominated by browsers and mixed feeders. Such a fauna, especially that with a large proportion of very large browsers (i.e., three species of giraffids, several large bovids and suids, chalicotheres, Ceratotherium, deinotheres), has no modern analogs, because there are no present-day ecosystems, beyond tropical forests, that have such a diverse guild of browsing herbivores. There is some evidence to suggest that conditions became slightly drier, with a greater proportion of grassland and open woodland, in the upper part of the Upper Laetolil Beds above Tuff 7. The evidence from the fossil mammals consistently points to a major shift in the Upper Ndolanya Beds to an ecosystem dominated by grassland. Further important evidence about the paleoecology is provided by the fossil gastropods (Tattersfield 2011). These indicate an abundance of woodland habitats throughout the Upper Laetolil Beds, but again they suggest that conditions became somewhat drier above Tuff 7. The gastropods from the Upper Ndolanya Beds, in contrast to the evidence from the fossil mammals, indicate that more mesic conditions prevailed, with extensive woodlands, similar to the paleoecology from the lower part of the Upper Laetolil Beds, which were the most mesic part of the sequence. A similar conclusion can be inferred from the oxygen isotope data from ostrich eggshell, which suggests that more mesic conditions were present in the Upper Ndolanya Beds. In addition, one of
T. Harrison
the main differences distinguishing the rodent community from the Upper Ndolanya Beds in comparison with the Upper Laetolil Beds, is the occurrence of Thryonomys (cane rat) (Denys 2011). The extant species of Thryonomys live in waterlogged valley bottoms and moist areas with reliable rainfall, where they specialize in feeding on coarse grasses and reeds (Kingdon 1997). Given that gastropods are highly sensitive indicators of the local ecology compared to most mammals, I am inclined to accept that the paleoecology of the Upper Ndolanya Beds was characterized by a greater extent of woodland than is indicated by the large mammal fauna. It is possible that the paleoecological signal derived from the large mammals is influenced by taphonomic factors (i.e., a bias towards larger-bodied ungulates) or that a significant part of the large mammal community may be transitory or migratory in nature, and therefore not reflective of the local ecology. The balance of the evidence would suggest that the paleoecology of the Upper Laetolil Beds was dominated by a mosaic of closed woodland, open woodland, shrubland and grassland. It was certainly more densely wooded than the modern-day Laetoli ecosystem, which is dominated by grassland and open woodland (Andrews et al. 2011). Water was probably more abundant during the rainy season, judging from the size and frequency of watercourses and small-scale fluvial deposits, but the region would have been dry for most of the year, except for the possible occurrence of permanent springs along the margin of the Eyasi Plateau. The paleoecology of the Lower Laetolil Beds was probably quite similar to that of the Upper Laetolil Beds. There is evidence, however, of semi-permanent bodies of water, but generally the inferred ecology is one of a dry woodland and bushland, possibly representative of an ecosystem that was disturbed by heavy inundations of volcanic ashes. The paleoecological reconstruction of the Upper Ndolanya Beds is more problematic because of the conflicting evidence derived from different proxies. However, it is very likely that conditions became drier than in the Upper Laetolil Beds, with a greater proportion of grassland, but that closed and open woodlands were still a significant part of the ecosystem. Acknowledgements A special thanks to all of the dedicated and resourceful team members who participated in the expeditions to Laetoli that contributed to the recovery of the material discussed and analyzed here. This volume and its companion would not have been possible without them. I would especially like to single out the following individuals who were critical to the success of the field project: Amandus Kweka, Michael L. Mbago, Charles P. Msuya, Simon Odunga, Al Deino, Carl Swisher, Peter Ditchfield, Godwin Mollel, Lindsay McHenry, Craig Feibel, Moses Lilombero, Simon Mataro, Denise Su, Peter Andrews, Terri Harrison and Bill Sanders. I thank all of the authors for their excellent contributions to this volume. For those who got their chapters in on time I am especially grateful; to those who were late with their submission, I hope that I am forgiven for the persistent nagging. To the senior physical anthropologist that accused me of doing stamp collecting rather
1 Hominins and Associated Fauna than science, I will let the content of this volume speak for itself. I thank the Tanzania Commission for Science and Technology and the Unit of Antiquities in Dar es Salaam for permission to conduct research in Tanzania. Special thanks go to Paul Msemwa (Director), Amandus Kweka and all of the curators and staff at the National Museum of Tanzania in Dar es Salaam for their support and assistance. Fieldwork at Laetoli was supported by grants from the National Geographic Society, the Leakey Foundation, and the National Science Foundation (Grants BCS-0216683 and BCS-0309513). This chapter is dedicated to the memory of the late Norbert Kayombo (former Director General of the National Museums of Tanzania) for his unwavering support.
References Andrews, P. J. (1989). Palaeoecology of Laetoli. Journal of Human Evolution, 18, 173–181. Andrews, P. (2006). Taphonomic effects of faunal impoverishment and faunal mixing. Palaeogeography, Palaeoclimatology, Palaeoecology, 241, 572–589. Andrews, P., & Bamford, M. (2008). Past and present ecology of Laetoli, Tanzania. Journal of Human Evolution, 54, 78–98. Andrews, P. J., & Humphrey, L. (1999). African Miocene environments and the transition to early hominines. In T. Bromage & F. Schrenk (Eds.), African biogeography climate change and human evolution (pp. 282–315). Oxford: Oxford University Press. Andrews, P., Bamford, M., Njau, E.-F., & Leliyo, G. (2011). The ecology and biogeography of the Endulen-Laetoli area in northern Tanzania. In T. Harrison (Ed.), Paleontology and geology of Laetoli: Human evolution in context (Geology, geochronology, paleoecology, and paleoenvironment, Vol. 1, pp. 167–200). Dordrecht: Springer. Armour-Chelu, M., & Bernor, R. L. (2011). Equidae. In T. Harrison (Ed.), Paleontology and geology of Laetoli: Human evolution in context (Fossil hominins and the associated fauna, Vol. 2, pp. 295–326). Dordrecht: Springer. Bamford, M. (2011a). Fossil leaves, fruits and seeds. In T. Harrison (Ed.), Paleontology and geology of Laetoli: Human evolution in context (Geology, geochronology, paleoecology and paleoenvironment, Vol. 1, pp. 235–252). Dordrecht: Springer. Bamford, M. (2011b). Fossil woods. In T. Harrison (Ed.), Paleontology and geology of Laetoli: Human evolution in context (Geology, geochronology, paleoecology and paleoenvironment, Vol. 1, pp. 217– 233). Dordrecht: Springer. Bishop, L. C. (2011). Suidae. In T. Harrison (Ed.), Paleontology and geology of Laetoli: Human evolution in context (Fossil hominins and the associated fauna, Vol. 2, pp. 327–337). Dordrecht: Springer. Bishop, L. C., Plummer, T. W., Hertel, F., & Kovarovic, K. (2011). Paleoenvironments of Laetoli, Tanzania as determined by antelope habitat preferences. In T. Harrison (Ed.), Paleontology and geology of Laetoli: Human evolution in context (Geology, geochronology, paleoecology and paleoenvironment, Vol. 1, pp. 355–366). Dordrecht: Springer. Cerling, T. E. (1992). Development of grasslands and savannas in East Africa during the Neogene. Palaeogeography, Palaeoclimatology, Palaeoecology, 97, 241–247. Deino, A. (2011). 40Ar/39Ar dating of Laetoli, Tanzania. In T. Harrison (Ed.), Paleontology and geology of Laetoli: Human evolution in context (Geology, geochronology, paleoecology and paleoenvironment, Vol. 1, pp. 77–97). Dordrecht: Springer. Denys, C. (2011). Rodents. In T. Harrison (Ed.), Paleontology and geology of Laetoli: Human evolution in context (Fossil hominins and the associated fauna, Vol. 2, pp. 15–53). Dordrecht: Springer. Ditchfield, P., & Harrison, T. (2011). Sedimentology, lithostratigraphy and depositional history of the Laetoli area. In T. Harrison (Ed.),
13 Paleontology and geology of Laetoli: Human evolution in context (Geology, geochronology, paleoecology and paleoenvironment, Vol. 1, pp. 47–76). Dordrecht: Springer. Gentry, A. W. (2011). Bovidae. In T. Harrison (Ed.), Paleontology and geology of Laetoli: Human evolution in context (Fossil hominins and the associated fauna, Vol. 2, pp. 363–465). Dordrecht: Springer. Harris, J. M. (1987). Summary. In M. D. Leakey & J. M. Harris (Eds.), Laetoli: A Pliocene site in northern Tanzania (pp. 524–531). Oxford: Clarendon. Harrison, T. (2005). Fossil bird eggs from Laetoli, Tanzania: Their taxonomic and paleoecological implications. Journal of African Earth Sciences, 41, 289–302. Harrison, T. (Ed.). (2011a). Paleontology and geology of Laetoli: Human evolution in context (Geology, geochronology, paleoecology and paleoenvironment, Vol. 1). Dordrecht: Springer. Harrison, T. (2011b). Hominins from the Upper Laetolil and Upper Ndolanya Beds, Laetoli. In T. Harrison (Ed.), Paleontology and geology of Laetoli: Human evolution in context (Fossil hominins and the associated fauna, Vol. 2, pp. 141–188). Dordrecht: Springer. Harrison, T. (2011c). Cercopithecids (Cercopithecidae, Primates). In T. Harrison (Ed.), Paleontology and geology of Laetoli: Human evolution in context (Fossil hominins and the associated fauna, Vol. 2, pp. 83–139). Dordrecht: Springer. Harrison, T. (2011d). Galagidae (Lorisoidea, Primates). In T. Harrison (Ed.), Paleontology and geology of Laetoli: Human evolution in context (Fossil hominins and the associated fauna, Vol. 2, pp. 75–81). Dordrecht: Springer. Harrison, T. (2011e). Laetoli revisited: Renewed paleontological and geological investigations at localities on the Eyasi Plateau in northern Tanzania. In T. Harrison (Ed.), Paleontology and geology of Laetoli: Human evolution in context (Geology, geochronology, paleoecology and paleoenvironment, Vol. 1, pp. 1–15). Dordrecht: Springer. Harrison, T. (2011f). Coprolites: Taphonomic and paleoecological implications. In T. Harrison (Ed.), Paleontology and geology of Laetoli: Human evolution in context (Geology, geochronology paleoecology and paleoenvironment, Vol. 1, pp. 279–292). Dordrecht: Springer. Harrison, T., & Kweka, A. (2011). Paleontological localities on the Eyasi Plateau, including Laetoli. In T. Harrison (Ed.), Paleontology and geology of Laetoli: Human evolution in context (Geology, geochronology, paleoecology, and paleoenvironment, Vol. 1, pp. 17–45). Dordrecht: Springer. Harrison, T., & Msuya, C. P. (2005). Fossil struthionid eggshells from Laetoli, Tanzania: Their taxonomic and biostratigraphic significance. Journal of African Earth Sciences, 41, 303–315. Hay, R. L. (1987). Geology of the Laetoli area. In M. D. Leakey & J. M. Harris (Eds.), Laetoli: A Pliocene site in northern Tanzania (pp. 23–47). Oxford: Clarendon. Hernesniemi, E., Giaourtsakis, I. X., Evans, A. R., & Fortelius, M. (2011). Rhinocerotidae. In T. Harrison (Ed.), Paleontology and geology of Laetoli: Human evolution in context (Fossil hominins and the associated fauna, Vol. 2, pp. 275–293). Dordrecht: Springer. Kaiser, T. M. (2011). Feeding ecology and niche partitioning of the Laetoli ungulate faunas. In T. Harrison (Ed.), Paleontology and geology of Laetoli: Human evolution in context (Geology, geochronology, paleoecology and paleoenvironment, Vol. 1, pp. 329–354). Dordrecht: Springer. Kingdon, J. (1997). The Kingdon field guide to African mammals. San Diego: Academic Press. Kingston, J. (2011). Stable isotopic analyses of Laetoli fossil herbivores. In T. Harrison (Ed.), Paleontology and geology of Laetoli: Human evolution in context (Geology, geochronology, paleoecology and paleoenvironment, Vol. 1, pp. 293–328). Dordrecht: Springer. Kingston, J., & Harrison, T. (2007). Isotopic dietary reconstructions of Pliocene herbivores at Laetoli: Implications for early hominin
14 paleoecology. Palaeogeography, Palaeoclimatology, Palaeoecology, 243, 272–306. Kitching, I. J., & Sadler, S. (2011). Lepidoptera, Insecta. In T. Harrison (Ed.), Paleontology and geology of Laetoli: Human evolution in context (Fossil hominins and the associated fauna, Vol. 2, pp. 549–554). Dordrecht: Springer. Kovarovic, K. (2004). Bovids as palaeoenvironmental indicators. An ecomorphological analysis of bovid postcranial remains from Laetoli, Tanzania. Ph.D. dissertation, University of London, London. Kovarovic, K., & Andrews, P. (2007). Bovid postcranial ecomorphological survey of the Laetoli paleoenvironment. Journal of Human Evolution, 52, 663–680. Kovarovic, K., & Andrews, P. (2011). Environmental change within the Laetoli fossiliferous sequence: Vegetation catenas and bovid ecomorphology. In T. Harrison (Ed.), Paleontology and geology of Laetoli: Human evolution in context (Geology, geochronology, paleoecology and paleoenvironment, Vol. 1, pp. 367–380). Dordrecht: Springer. Kovarovic, K., Andrews, P., & Aiello, L. (2002). An ecological diversity analysis of the Upper Ndolanya Beds, Laetoli, Tanzania. Journal of Human Evolution, 43, 395–418. Krell, F.-T., & Schawaller, W. (2011). Beetles (Insecta: Coleoptera). In T. Harrison (Ed.), Paleontology and geology of Laetoli: Human evolution in context (Fossil hominins and the associated fauna, Vol. 2, pp. 535–548). Dordrecht: Springer. Leakey, M. D., & Harris, J. M. (Eds.). (1987). Laetoli: A Pliocene site in northern Tanzania. Oxford: Clarendon. Louchart, A. (2011). Aves. In T. Harrison (Ed.), Paleontology and geology of Laetoli: Human evolution in context (Fossil hominins and the associated fauna, Vol. 2, pp. 505–533). Dordrecht: Springer. Manega, P. (1993). Geochronology, geochemistry and isotopic study of the Plio-Pleistocene hominid sites and the Ngorongora volcanic highlands in northern Tanzania. Ph.D. dissertation, University of Colorado at Boulder, Boulder. Musiba, C. M. (1999). Laetoli Pliocene paleoecology: A reanalysis via morphological and behavioral approaches. Ph.D. dissertation, University of Chicago, Chicago. Musiba, V., Magori, C., Stoller, M., Stein, T., Branting, S., & Vogt, M. (2007). Taphonomy and paleoecological context of the Upper Laetolil Beds (Localities 8 and 9), Laetoli in northern Tanzania. In R. Bobe, Z. Alemseged, & A. K. Behrensmeyer (Eds.), Hominin environments in the East African Pliocene: An assessment of the faunal evidence (pp. 257–278). Dordrecht: Springer. Ndessokia, P. N. S. (1990). The mammalian fauna and archaeology of the Ndolanya and Olpiro Beds, Laetoli, Tanzania. Ph.D. dissertation, University of California, Berkeley. Peters, C. R., Blumenschine, R. J., Hay, R. L., Livingstone, D. A., Marean, C. W., Harrison, T., Armour-Chelu, M., Andrews, P., Bernor, R. L., Bonnefille, R., & Werdelin, L. (2008). Paleoecology of the Serengeti-Mara ecosystem. In A. R. E. Sinclair, C. Packer, S. A. R. Mduma, & J. M. Fryxell (Eds.), Serengeti III: Human impacts on ecosystem dynamics (pp. 47–94). Chicago: University of Chicago Press.
T. Harrison Rage, J. C., & Bailon, S. (2011). Amphibia and Squamata. In T. Harrison (Ed.), Paleontology and geology of Laetoli: Human evolution in context (Fossil hominins and the associated fauna, Vol. 2, pp. 467–478). Dordrecht: Springer. Reed, D. (2011). Serengeti micromammal communities and the paleoecology of Laetoli, Tanzania. In T. Harrison (Ed.), Paleontology and geology of Laetoli: Human evolution in context (Geology, geochronology, paleoecology and paleoenvironment, Vol. 1, pp. 253–263). Dordrecht: Springer. Reed, D., & Denys, C. (2011). The taphonomy and paleoenvironmental implications of the Laetoli micromammals. In T. Harrison (Ed.), Paleontology and geology of Laetoli: Human evolution in context (Geology, geochronology, paleoecology and paleoenvironment, Vol. 1, pp. 265–278). Dordrecht: Springer. Robinson, C. (2011). Giraffidae. In T. Harrison (Ed.), Paleontology and geology of Laetoli: Human evolution in context (Fossil hominins and the associated fauna, Vol. 2, pp. 339–362). Dordrecht: Springer. Rossouw, L., & Scott, L. (2011). Phytoliths and pollen, the microscopic plant remains in Pliocene volcanic sediments around Laetoli, Tanzania. In T. Harrison (Ed.), Paleontology and geology of Laetoli: Human evolution in context (Geology, geochronology, paleoecology and paleoenvironment, Vol. 1, pp. 201–215). Dordrecht: Springer. Sanders, W. L. (2011). Proboscidea. In T. Harrison (Ed.), Paleontology and geology of Laetoli: Human evolution in context (Fossil hominins and the associated fauna, Vol. 2, pp. 189–232). Dordrecht: Springer. Su, D. (2005). The paleoecology of Laetoli, Tanzania: Evidence from the mammalian fauna of the Upper Laetolil Beds. Ph.D. dissertation, New York University, New York. Su, D. F. (2011). Large mammal evidence for the paleoenvironment of the Upper Laetolil and Upper Ndolanya Beds of Laetoli, Tanzania. In T. Harrison (Ed.), Paleontology and geology of Laetoli: Human evolution in context (Geology, geochronology, paleoecology and paleoenvironment, Vol. 1, pp. 381–392). Dordrecht: Springer. Su, D. F., & Harrison, T. (2007). The paleoecology of the Upper Laetolil Beds at Laetoli: A reconsideration of the large mammal evidence. In R. Bobe, Z. Alemseged, & A. K. Behrensmeyer (Eds.), Hominin environments in the East African Pliocene: An assessment of the faunal evidence (pp. 279–313). Dordrecht: Springer. Su, D. F., & Harrison, T. (2008). Ecological implications of the relative rarity of fossil hominins at Laetoli. Journal of Human Evolution, 55, 672–681. Tattersfield, P. (2011). Gastropoda. In T. Harrison (Ed.), Paleontology and geology of Laetoli: Human evolution in context (Fossil hominins and the associated fauna, Vol. 2, pp. 567–587). Dordrecht: Springer. Werdelin, L., & Dehghani, R. (2011). Carnivora. In T. Harrison (Ed.), Paleontology and geology of Laetoli: Human evolution in context (Fossil hominins and the associated fauna, Vol. 2, pp. 189–232). Dordrecht: Springer. Winkler, A., & Tomida, Y. (2011). The lower third premolar of Serengetilagus praecapensis (Mammalia: Lagomorpha: Leporidae) from Laetoli, Tanzania. In T. Harrison (Ed.), Paleontology and geology of Laetoli: Human evolution in context (Fossil hominins and the associated fauna, Vol. 2, pp. 55–66). Dordrecht: Springer.
Chapter 2
Rodents Christiane Denys
Abstract New rodent specimens collected at Laetoli between 1998 and 2005 are described here. The material allows an updating and refinement of the previously published taxonomic lists, especially those for the Lower Laetolil Beds and the Upper Ndolanya Beds. The increased number of well-preserved cranial specimens allows the description of several new species and a better appreciation of the size and morphology of some Laetoli taxa compared to their southern and eastern African counterparts. This is especially the case for Saccostomus, for which the fossil record has recently been much improved. The new species described here include a small sciurid, two Gerbillinae, and a thryonomyid. Some species are newly recognized at certain localities, and Aethomys and Petromus are recorded for the first time at Laetoli. The distribution and stratigraphic range for Pedetes laetoliensis is extended, and it is now recorded in the Upper Ndolanya Beds. Similarly, Xerus janenschi is now identified in the Laetolil Beds. As in the previous study of the Laetoli rodents, important differences in species composition and diversity between the Upper Laetolil Beds and the Upper Ndolanya Beds are confirmed. These probably reflect differences in landscape. Compared to other Pliocene assemblages, the Laetolil Beds are characterized by a very unusual diversity of sciurids and the dominance of Saccostomus and Pedetes, but otherwise they compare well with other East African Mio-Pliocene rodent assemblages, such as those from the Omo Valley and Lemudong’o. The Laetoli assemblages are distinct from those of Lukeino, Chorora and Harasib 3, but could belong to the same faunal unit as Ibole (Manonga Valley). They also differ in some respects from those from Hadar and Pliocene South African sites. Few species are shared in common between the Laetolil Beds and Upper Ndolanya Beds, but it is uncertain whether this turnover is due to taphonomic or paleoclimatic factors. This contribution highlights the importance of Laetoli for
C. Denys (*) Department of Systematics and Evolution – CP51, UMR7205 CNRS: Origine structure & évolution de la Biodiversité, MNHN, 55 rue Buffon, 75005, Paris, France e-mail:
[email protected] understanding rodent evolution, as well as for its geographic position at the crossroads between East and South Africa. Keywords Mammalia • Rodentia • East Africa • Pliocene • Pleistocene • Taxonomy
Introduction In Africa, small mammals represent about 80% of the modern biodiversity, and rodents alone constitute about the half of it. Their role as primary consumers and forest regenerators make them important in ecosystems, and they are considered good indicators of habitat. Due to their relatively small size, fossil rodents occur only in localized bone concentrations, and among the Pliocene sites of Africa there are few rodent faunas known. The Laetoli rodents were initially collected during the 1938–1939 Kohl-Larsen expedition to the southern Serengeti, which formed the basis for Dietrich’s (1942) initial taxonomic study. Subsequent collections by Mary Leakey (1974–1979) allowed a better documentation of rodent paleodiversity (Denys 1987a; Davies 1987) and situated the faunas in a wellconstrained geochronological and stratigraphic context for the first time. This led to an improved knowledge of rodent evolution during the Plio-Pleistocene of East Africa, including a better appreciation of their relationships with South African faunas (Denys 1999; Denys et al. 2003; Winkler et al. 2010). Due to the peculiar sedimentary nature of the site, Laetoli is characterized by remarkably well-preserved material, which allows the description of cranial and postcranial characteristics of the rodents. Laetoli provides records of the first appearance data (FAD) of several rodent genera and, being located at the southern end of the Rift Valley, it is biogeographically important. Moreover, rodents are known both from the Laetolil Beds (lower and upper units) and the Upper Ndolanya Beds, which allows biostratigraphical comparisons between the main stratigraphic units. We present here the results of a systematic study of new fossil rodent material recovered by Terry Harrison’s teams during
T. Harrison (ed.), Paleontology and Geology of Laetoli: Human Evolution in Context. Volume 2: Fossil Hominins and the Associated Fauna, Vertebrate Paleobiology and Paleoanthropology, DOI 10.1007/978-90-481-9962-4_2, © Springer Science+Business Media B.V. 2011
15
16
the 1998–2005 field seasons at Laetoli. The study includes the description of new taxa and a reinterpretation of the evolutionary relationships of the fossil rodents from Laetoli.
Material and Methods Specimens were examined and illustrated using a Wild Microscope fitted with a camera lucida. Cranial and dental dimensions were measured with Mitutoyo calipers (0.01 mm precision). Some specimens were prepared by R. Vacant (Palaeontology Laboratory at the MNHN) and by the author. SEM images of the teeth were taken by C. Chancogne-Weber with a JEOL 45 at the Palaeontology Laboratory. Univariate statistics were performed using XLSTAT Software version 9 (Addinsoft). Comparisons were made with the following reference mammal collections: Paris, France (MNHN); Natural History Museum, London, England (NHM); Zoologische Museum, Berlin, Germany (ZMB); Zoologishe Museum für Naturkunde, Bonn, Germany (ZFMK); Durban Science Museum, South Africa (DM); Namibian Museum, Windhoek, Namibia (NM). Tooth nomenclature follows Denys (1987a), and rodent taxonomy follows that of Wilson and Reeder (2005).
C. Denys
Springhares are quite numerous at Laetoli, with well-preserved skeletal material. The specimens collected by Mary Leakey led Davies (1987) to describe a new species. Among the diagnostic characters were its small size, enlarged infraorbital foramen and the absence of cusps on the molars (Fig. 2.1). The original type description did not list the provenance of the specimens, but Davies (personal communication) listed 35 individuals of Pedetes occurring at Locs. 1, 2, 3, 4, 5, 6, 7, 8, 9, 9N, 9S, 10, 10W, 10E, 11, 13, 14, 15, 16, 19, 21 and 22. Davies (1987) mentions the occurrence of Pedetes cf. surdaster from the Late Pleistocene Upper Ngaloba Beds at Loc. 2, but none from the Upper Ndolanya Beds. However, Harris (1987) lists the species as occurring in the Upper Ndolanya Beds. Here, 75 additional specimens add to the number of localities at which Pedetes occurs (see Appendix 2.1). The new remains come from Locs.1, 2, 4, 5, 6, 8, 9, 10E, 11, 13, 15, 21 and 22, and are derived from all horizons throughout Table 2.1 Upper and lower toothrow length (mm) for the new Laetolil Beds specimens of Pedetes laetoliensis Davies, 1987, compared with the dimensions of the holotype (after Davies 1987) and representatives of the two extant species Specimen
P/4-M/3 12.38 12.94 13.04 13.15
Suborder Anomaluromorpha Bugge, 1974 Family Pedetidae Gray, 1825 Pedetes laetoliensis Davies, 1987 (Fig. 2.1, Table 2.1)
EP 1089/05 EP 714/00 EP 1509/98 EP 1235/98 EP 2914/00 Holotype P. capensis P. surdaster P. capensis N= 4
Fig. 2.1 New specimens of Pedetes from Laetoli. (a) right maxilla with DP4-M3/ of P. laetoliensis (EP 1994/00, Loc. 5, Upper Laetolil Beds); (b) right mandible with DP/4-M/3 of P. laetoliensis (EP 1867/00,
Loc. 2, Upper Laetolil Beds); (c) mandible of Pedetes sp. with DP/4-M/2 (EP 2196/00, Loc. 7E, Upper Ndolanya Beds). Scale bar in mm
Systematics
14.05 17.9 18.5 Mean 17.26 Range 16.67–17.38
P4/-M3/
13.02 13.5 19.1 18.0 Mean 17.42 Range 16.60–18.94
2 Rodents
the Upper Laetolil Beds. The dimensions of the upper and lower toothrows of the new specimens are close to those of the type series, but they display a great range of variability (Table 2.1). This may be due to the difficulty in measuring some isolated molars that have convex crowns and because the occlusal surface of hypsodont molars changes in dimensions during the course of wear. The shape of the molars is similar to the previously recovered material described by Davies (1987: fig. 6.29, p. 176) (Fig. 2.1). The molars are characterized by bilobate crowns of nearly equal size and proportions, which makes identification of serial position difficult. They all have high crowns and flat occlusal surfaces. No traces of cusps are visible. The only other extinct species of Pedetes, P. gracilis, comes from Taung (Broom 1934: fig. 5, p. 476). Pedetes gracilis has a longer molar row (12 mm) and is very similar to the modern Pedetes caffer. According to Broom (1934), the differences between the species are the smaller size of the fossil teeth, and the plates of the infolded enamel are nearly parallel and less deeply folded than those in modern P. caffer. Molars of P. laetoliensis display deep folds and are not fully parallel in comparison to P. gracilis. A pedetid indet. is mentioned briefly as occurring at Harasib, and probably constitutes a new genus of the family (Mein et al. 2000a). In addition, a single incisor from Lukeino (Mein and Pickford 2006) is attributed to an indeterminate Pedetidae.
17
Genus Paraxerus Forsyth Major, 1893 This taxon is characterized by a short zygomatic plate, complex upper molars with three clear re-entrant folds, lower teeth with central depression non-isolated and well-developed ectolophid. When the lower molars have strongly marked cusps and non-flattened crowns during wear, one can attribute the molars to Paraxerus rather than to Funisciurus. Both genera have a P3/. Paraxerus meini sp. nov. (Fig. 2.2) Holotype: EP 2816/00, left mandible with P/4-M/3 (Fig. 2.2). Type locality: Laetoli Loc. 5, Upper Laetolil Beds between Tuffs 3 and 5, Tanzania.
Pedetes sp. Only one specimen has been recovered from the Upper Ndolanya Beds at Loc. 7E during renewed fieldwork, while Davies (personal communication) recorded its presence at Loc. 14. It is represented by a mandible with DP/4-M/2 (EP 2196/00) in a poor state of preservation (Fig. 2.1). The length of the DP/4-M/2 reaches 10.06 mm in EP 2196/00, which, based on its small size, indicates the possible presence of P. laetoliensis in the Upper Ndolanya Beds. Up to now no Pedetes has been recovered from the Upper Ndolanya Beds at Loc. 18. The molars display no link between the two lobes of the molars, and the first lobe of the P4 shows two wellindividualized and oblique cusps, which is considered a juvenile feature. Family Sciuridae Fischer de Waldheim, 1817 Sciurid remains are quite abundant at Laetoli. From the Laetolil Beds three different taxa of sciurid were recognized by Denys (1987a), a small Paraxerus sp. (Locs. 11 and 12), a larger Xerus sp. (Loc. 9S), and Xerus cf. janenschi (Loc. 2). The Upper Ndolanya Beds at Locs. 7E and 18 have yielded well-preserved remains of Xerus janenschi. Newly recovered cranial material allows us to refine the taxonomy of the Laetoli squirrels, which can be distinguished on their molar row size and dental criteria.
Fig. 2.2 Paraxerus meini sp. nov. upper and lower molars. Top, EP 881/03 (paratype) (Loc. 10E, Upper Laetolil Beds), left maxillary fragment with P4/M1; below, EP 1250/03 (Loc.7E, Upper Ndolanya Beds), right mandible with P/4-M/2 and EP 2816/00 (holotype) (Loc. 5, Upper Laetolil Beds)
18
C. Denys
Age and Horizon: Mid-Pliocene, Upper Laetolil Beds (between Tuffs 3 and 7) and Upper Ndolanya Beds. Paratypes: EP 881/03 (Loc. 10E), maxillary fragment with P4-M1 (Fig. 2.2). EP 2815/00 (Loc. 5), right mandibular fragment with P/4-M/2. EP 1000/01 (Loc. 11), mandible with P/4-M/3. EP 4152/00 (Loc. 8), right and left hemi-mandibles with M/1-3. EP 1250/03 (Loc. 7E), right mandible with P/4-M/2. Referred material from Laetolil Beds (previously identified as Paraxerus sp. indet. by Denys 1987a): LAET 76-4121A, lower P/4; LAET 74-304, upper P4-M3/ (figured in Denys 1987a, plate 6.2-1 p. 123); LAET 76-4178, lower P/4-M/1; LAET 76-4170, right mandible fragment with P/4- M/3 (figured in Denys 1987a: plate 6.2-2, p. 123). Distribution: Localities 5, 8, 10E, 11, and 12 of the Upper Laetolil Beds, and Loc. 7E of the Upper Ndolanya Beds. Repository: National Museum of Tanzania, Dar es Salaam. Etymology: Named in honor of Pierre Mein, who has described many new rodent species from the Miocene of Europe and Africa. Measurements: Tables 2.2 and 2.3. Diagnosis: One of the smallest species of the genus compared to modern Paraxerus. Smaller than extant P. ochraceus, which is the smallest East African species, but larger than P. boehmi from Central Africa. Bunodont, with many supplementary cusplets in all parts of the molars, more than in P. ochraceus. Less bunodont than modern P. ochraceus, P. palliatus, P. flavovittis, P. cepapi. Characterized by lower molars with a very rectilinear, long ectolophid associated with a mesoconid on M/1-2. Characterized by a transverse entolophid connected directly to the anterior part of the hypoconid. Differs from Heteroxerus karsticus in its smaller size, and the absence of a direct link between the entoconid and hypoconulid. Differs from P. ochraceus from the Omo in the larger size of the lower molars. Description and comparisons: During Mary Leakey’s expeditions of 1975–1976 Paraxerus was recovered only from
Locs. 11 and 12, and was represented by only three mandibular fragments and one maxillary fragment. Here we add and figure additional material from Locs. 5, 8 and 7E. This rare squirrel at Laetoli is represented by a few mandibles and incomplete maxillae, but no other cranial fragments. The P3/ occurs in all specimens, but only an alveolus is found, so that the morphology of the tooth cannot be described. All the preserved upper molars are heavily worn (Fig. 2.2; see Plate 6.2 in Denys 1987a), but one can distinguish an anteroloph and a small posteroloph on P4/. A paraloph and metaloph are visible with the development of a faint metaconule on the metaloph. The hypocone is hardly visible and no mesostyle is seen in specimen LAET 74-304, but one is found in specimen EP 881/03. On the upper M1 and M2 there are two well-developed parallel lophs. A small anteroloph exists, but the conules are not visible due to wear. The M3/ is present only on specimen LAET 74-304, but it is worn. It has a triangular shape and it is smaller than M1-2/. Two lophs are visible on M3/; the metaloph being reduced to a cusp in comparison to the protoloph. On P/4 the cusps are bunodont and the protoconid and metaconid are nearly the same height. The two cusps are united by a small crest issuing from the posterior part of the protoconid. There is an ectolophid linking the protoconid to the hypoconid, and a small posterolophid. No anterolophid is observed on P/4. On M/1-2 there is a small anterolophid and posterolophid with supplementary cusps (anteroconulid and hypoconulid). The ectolophid is well developed and longitudinal, with a slight mesoconid on M/1-2. The entolophid is well-developed and connects the hypoconid to the entoconid transversely with a very rectilinear crest. On the entolophid of M/1-2 there is one or two supplementary cusps. On M/2 the entolophid is smaller than on M/1 and the anteroconulid and hypoconulid are less visible. On M/3, which is narrow and elongated, the same structures are visible and the cusps are still distinguishable. The anteroconulid is low and small. The hypoconulid and posterolophid is absent on M/3. The
Table 2.2 Upper (UPTR) and lower (LTR) toothrow lengths (mm) for Paraxerus meini nov. sp. compared to modern Paraxerus species Species
Country/site
Laetoli (this work; Denys 1987a) Tanzania Tanzania Zimbabwe South Africa P. boehmi Uganda DR Congo P. ochraceus Kenya Somalia Tanzania N number of specimens, SD standard deviation P. meini P. flavovittis P. cepapi
N
UPTR mean
SD
Range
N
LTR mean
SD
Range
2 12 10
6.63 7.60 7.93
0.53 0.316 0.52
6.25–7.00 7.03–8.18 7.08–8.43
4 14 10
6.88 7.41 7.66
0.69 0.213 0.55
5.96–7.58 7.02–7.76 6.8 – 8.29
2
6.13
5.48–6.79
2
5.60
4
7.09
6.32–7.72
4
6.88
0.57
5.57–5.64 0.49
6.25–6.97
2 Rodents
19
Table 2.3 Fossil and modern Paraxerus spp. lower P4 and molar dimensions (mm) P/4L
P/4W
P. meini LAET 4121A LAET 4178 LAET 4170 EP 2815/00 EP 2816/00 EP 4152/00 EP 1000/01 Mean
1.95 1.78 1.70 1.52 1.52 1.76 1.67 1.70
1.65 1.50
P. ochraceus 26.5.12.50 P. flavovittis 2007-1236 Paraxerus sp. KNM-NK 44920 KNM-KP 46313
M/1L
M/1W
M/2L
M/2W
M/3L
M/3W
1.74 1.68 1.76 1.57 1.67 1.67 1.68
1.82 1.71 1.71 1.81 1.71 1.75
1.75 1.90 1.71 1.90 1.76 1.80
2.02
1.70
1.38 1.29 1.62 1.52 1.49
1.85 1.75 1.90 1.71 1.76 1.67 1.77
2.00 2.10 1.71 1.96
1.62 1.81 1.71 1.71
1.57
1.29
1.62
1.48
1.62
1.66
1.86
1.48
1.52
1.52
1.57
1.62
1.72
1.86
2.05
1.62
2.08
2.0 2.0
2.2
2.0
2.6
3.0
2.5
P. cepapi DM521
1.81
1.76
1.91
1.91
1.95
2.05
1.91
1.86
P. ochraceus Omo B
1.5
1.45
1.765
1.61–1.86
1.8
1.7–1.87
2.16 1.85–2.3
1.97 1.8–2.04
H. karsticus Mean 1.44 1.45 1.81 1.82 1.99 1.88 2.06 1.85 Minimum 1.34 1.36 1.66 1.73 1.77 1.70 1.89 1.72 Maximum 1.54 1.55 1.97 1.92 2.24 2.06 2.23 2.03 Standard deviation 0.074 0.059 0.083 0.056 0.11 0.093 0.1 0.103 L length, W width Sources: P. ochraceus, Omo Member B, Wesselman (1984); H. karsticus, Mein et al. (2000a); Paraxerus sp., Kanapoi and Lemudong’o, Manthi (2006, 2007). Modern species (P. ochraceus, P. flavovittis and P. cepapi) from museum collections
entoconid is small and oblique, delimiting the distal border of the molar. Comparisons of molar size with various modern East and South African Paraxerus species shows that Paraxerus meini nov. sp. clearly has a smaller toothrow length compared to modern P. flavovittis, P. cepapi and P. ochraceus, but larger than P. boehmi. There is marked individual variability of molar size in the modern species (Tables 2.2 and 2.3). Comparison of the morphological features of the molars with modern P. ochraceus shows that P. meini shares well-developed lophs on the upper molars, the presence of a metaconule on P4/, and no hypocone on M3/. Lophs are less well developed in P. palliatus and P. flavovittis, especially the ectolophid. On the lower molars one can see the anteroconulid and posteroconulid on M/1-2 of P. palliatus and P. ochraceus, and the M/3 is narrow and lacks a hypoconulid and discrete entoconid. It seems that P. meini can be distinguished from P. palliatus and P. flavovittis in having numerous supplementary cusplets, less bunodont molars, and a better-developed ectolophid. Compared to P. ochraceus there are fewer supplementary cusplets and a more rectilinear ectolophid
with a mesoconid. Paraxerus meini differs from P. flavovittis in the presence of a hypoconulid and entoconulid on M/1 and M/2. From a morphological point of view the molars of P. meini display some similarities with P. ochraceus from Tanzania, being characterized by the development of numerous cusplets, but they are expressed to a greater extent in P. meini. This group is characterized by marked molar size and shape variability, but cusp variability is not well known. From the MNHN, NHM, ZMB and DM voucher specimens examined, P. meini has upper and lower toothrows intermediate in size between P. boehmi and P. ochraceus (Table 2.3). The fossil record for this taxon is poorly known. Only an isolated tooth (left P/4) of Paraxerus sp. has been discovered at Lemudong’o (~6 Ma), which displays some similarity with P. palliatus (Manthi 2007). Compared to P. meini, the Lemudong’o P/4 has the same bunodont pattern with the two anterior cusps well separated. Based on the published images of the Lemudong’o specimen there is no ectolophid, contrary to P. meini, and no evidence of a posterolophid. Another lower molar attributed to Paraxerus sp. was described from Tabarin (4.5–4.4 Ma) (Winkler 2002). Manthi (2006) mentions a single mandible of Paraxerus sp. from Nzube’s mandible
20
site at Kanapoi, which has larger molars, similar to the specimen from Lemudong’o (Table 2.3). From Omo Members B, C and F, Wesselman (1984) described some molars that he attributed to modern P. ochraceus. They display the same pattern as P. meini: longitudinal rectilinear ectolophid, presence of an entolophid on M/1-2, and existence of an anterolophid or an anteroconulid. The difference between the Omo P. ochraceus and the modern species relates to the oblique disposition of the protoloph originating from the protocone in the fossil, while it is more transverse and originates from the back of the molar in the modern form. The specimens from Omo and Laetoli probably belong to the same lineage, and may be the ancestors of modern P. ochraceus. The P. meini specimens have larger molars compared to modern representatives of the genus (Table 2.3). The late Miocene site of Harasib in Namibia has yielded the remains of a sciurid that is attributed to the extinct European genus Heteroxerus (Mein et al. 2000a). The reasons why the Harasib squirrel is not attributed to Paraxerus is not well justified, except for the smaller size of the unicuspid P3/ in the Harasib material. However, there is extensive variability in the modern representatives, and such a character is not adequate to reject a close relationship between P. meini and H. karsticus. The hypocone is absent or small in H. karsticus and there is some variability described by the authors in the metaloph orientation and disposition. Heteroxerus karsticus has larger molars than P. meini. The figured holotype of H. karsticus displays a very longitudinal rectilinear ectolophid and there is an anteroconulid on M/1-2 as in P. meini. The entolophid is better developed and more transverse than in the Laetoli specimens, while it is very reduced or absent in H. karsticus (Mein et al. 2000a). When the entolophid is figured, as in Fig. 2.3, one observes that it is oblique and joins the
Fig. 2.3 Dorsal and ventral views of Xerus janenschi cranium, EP 219/04 from Loc. 15 (Upper Ndolanya Beds)
C. Denys
posterolophid midway along its length. The genus Heteroxerus was created by Stehlin and Schaub (1951) for the Miocene European H. hurzeleri based upon the existence of a direct link between the entoconid and hypoconulid, a feature that we do not find in Paraxerus meini or modern Paraxerus spp., but present on H. karsticus at Harasib (Mein et al. 2000a). Stehlin and Schaub (1951) also mentioned the existence of the little arm of the protoconid, which is also found in modern Xerus spp., but not in the Paraxerus we examined. Heteroxerus karsticus, as described by Mein et al. (2000a), also displays an anteroconulid on M/1-2 that is found in P. meini and in modern P. ochraceus and P. cepapi. The diagnostic characters provided by Mein et al. (2000a) indicate some differences between the two species and they probably represent distinct lineages. Further revisions of Heteroxerus and Paraxerus species composition and diagnoses are required to answer these questions. Genus Xerus Hemprich and Ehrenberg, 1833 Xerus janenschi Dietrich, 1942 (Figs. 2.3–2.6) The largest sciurid from Laetoli is found as a common taxon in the Upper Ndolanya Beds, but it is represented only by a single specimen from the Upper Laetolil Beds (Denys 1987a). Dietrich (1942) described it for the first time from Garusi, but the stratigraphic provenance and age is not known. Denys (1987a) recognized the same species from Locs. 18 and 7E from the Upper Ndolanya Beds.
Fig. 2.4 Scatter plot of modern and fossil Xerus specimens. LP4-M3/: upper tooth row length. LGT: total length of the cranium (axis scales in mm). ERY: modern Xerus erythropus from East and Central Africa. INAURIS: modern X. inauris from South Africa. JANENSCHI: Laetoli fossils, X. janenschi. PRINCEPS: modern Xerus princeps from southwest Africa. RUTILUS: modern X. rutilus from Ethiopia. SP: Laetolil Xerus sp. DAAMSI: Fossil Chad KB, X. daamsi
2 Rodents
Fig. 2.5 Box plots for the different modern and fossil Xerus spp. for different cranial measurements (LP4-M3/: Upper tooth row length. WNAS: Nasal width. LP/4-M/3: Lower tooth row length. LNAS: Nasal length. LGT: Total cranium length) in mm. ERY: modern Xerus erythropus from
21
East and Central Africa. INAURIS: modern X. inauris from South Africa. JANENSCHI: Laetoli fossils, X. janenschi. PRINCEPS: modern Xerus princeps from southwest Africa. RUTILUS: modern X. rutilus from Ethiopia. SP: Laetolil Xerus sp. DAAMSI: Fossil Chad KB, Xerus daamsi
22
C. Denys Table 2.4 Skull and molar dimensions (mm) in fossil and modern Xerus spp. Species
Fig. 2.6 Lower (left) and upper (right) toothrows of Xerus janenschi. EP 292/04 (mandible with P/4-M/2) from Loc. 18 (Upper Ndolanya Beds) and EP 2356/98 (maxilla with P4-M3/) from Loc. 18 (Upper Ndolanya Beds). Scale bar indicates 1 mm
Lacking P3/, no ectolophid and quite bunodont cusps, Xerus janenschi shares dental and cranial characters with X. daamsi (Denys et al. 2003) from the early Pliocene KB site in Chad and with X. erythropus from the Omo (Wesselman 1984). The absence of P3/ distinguishes the Laetoli fossils from the late Miocene Xerus sp. from Alayla Vertebrate Locality 2 in the Middle Awash of Ethiopia (Wesselman et al. 2009). A cladistic analysis comparing the Laetoli fossil to modern Xerini indicates that its closest affinities are with X. rutilus (Denys et al. 2003). The new Laetoli material increases the sample of X. janenschi by 37 specimens (see Appendix 2.2) and establishes its presence for the first time in the Upper Ndolanya Beds at Loc.15 and Silal Artum, as well as in the Upper Laetolil Beds at Loc. 9S. The new specimens display the same skull characteristics previously described for X. janenschi, including a short nasal, trace of three transbullae septa, rather bunodont molars, the absence of P3/, inflated tympanic bullae, and a wide braincase (Fig. 2.3). These characters allow X. janenschi to be grouped closest to the South African X. inauris and X. princeps. However, X. janenschi is also characterized by distinctive skull proportions (Table 2.4, Fig. 2.4). At an equivalent cranial
LGT
LNAS
WNAS
LP4/-M3/
LP/4-M/3
Xerus daamsi Chad (KB) 54.59 15.64 7.06 11.62 11.08 X. erythropus Mean 59.13 18.07 7.91 11.84 12.01 SD 3.26 1.21 0.6 0.73 0.46 Min 52.11 16.38 6.86 10.69 10.92 Max 65.06 20.33 8.99 13.27 12.68 X. rutilus Mean 51.07 13.26 6.74 9.39 9.65 SD 2.48 5.59 0.5 0.48 0.42 Min 45.44 12.86 6 8.27 8.71 Max 56.36 17.64 7.72 10.32 10.5 X. inauris Mean 55.65 17.35 8.44 10.97 11.89 SD 2.86 1.31 0.62 0.32 0.4 Min 51.91 15.47 7.72 10.58 11.14 Max 58.67 18.97 9.45 11.33 12.24 X. princeps Mean 58.5 19.89 7.74 11.16 12.01 SD 1.87 1.56 0.2 0.48 0.45 Min 56.6 17.82 7.44 10.57 11.52 Max 60.75 21.19 7.88 11.69 12.6 X. janenschi N 5 4 5 7 6 Mean 51.86 16.24 7.98 11.48 12.44 SD 0.99 0.65 0.28 0.24 0.21 Min 47.96 14.66 7 10.23 11.77 Max 53.4 17.83 8.51 12.28 13.12 Xerus sp. Berlin Gadj. 44.8 12.2 100, Laetoli N number of molars, SD standard deviation, min-max minimum and maximum values, LGT greatest length of the skull, LNAS and WNAS nasal length and width, LP4/-M3/ upper tooth row length, P/4-M/3, lower tooth row length. Modern X. inauris, X. princeps, X. rutilus, X. erythropus specimens have been measured in museum collections. Data for X. daamsi from Denys et al. (2003)
size to X. rutilus, X. janenschi has longer upper and lower molar rows. It is smaller than X. daamsi, X. princeps, X. inauris and X. erythropus. The nasals of Xerus janenschi are intermediate in length-width proportions between the smallest X. rutilus and X. daamsi and the other modern species, which are larger (Table 2.4, Fig. 2.5). There is great variability within this species in terms of size, but the dental morphology of the newly collected fossils is similar to the type material and displays a very bunodont pattern (Fig. 2.6). The new material confirms that X. janenschi has larger molars than X. daamsi from Chad and Xerus sp. from Kanapoi, and smaller molars than those of X. cf. inauris from Olduvai Bed I (Table 2.5). However, they fall within the lower end of the range of variability for the Olduvai and Omo samples (Table 2.5).
2 Rodents
23
Table 2.5 Tooth dimensions (N number of specimens, SD standard deviation) for fossil and modern Xerus spp. Laetoli UNB 1987 and this work relates to X. janenschi Length Width Tooth
Locality
N
Mean
SD
Range
N
Mean
SD
Range
P4/ or DP4/
Laetoli UNB 1987 4 3.13 0.4 2.49–3.77 4 2.65 0.62 1.66–3.63 Laetoli UNB this work 4 2.63 0.12 2.29–2.86 4 2.9 0.23 2.24–3.29 Olduvai 1 2.75 1 2.28 Omo F 1 2.55 1 2.98 KB-97-162 1 2.50 1 2.80 M1/ Laetoli UNB 1987 3 3.08 0.4 2.39–3.78 3 3.72 0.26 3.08–4.37 Laetoli UNB this work 4 2.88 0.06 2.95–3.67 4 3.31 0.15 2.95–3.57 Olduvai a 5 3.49 3.40–3.60 5 3.36 3.2–3.6 KB-97-162 1 2.86 1 2.86 M2/ Laetoli UNB 1987 2 3.08 0.28 2 3.55 Laetoli UNB this work 5 3.03 0.13 2.81–3.36 5 3.40 0.11 3–3.68 KB-97-162 1 2.76 1 3.05 M3/ Laetoli UNB 1987 2 2.83 2 3.10 Laetoli UNB this work 3 2.81 0.17 2.52–3.1 3 3.13 0.05 3.05–3.23 KB-97-162 1 2.76 1 3.05 P/4 Laetoli UNB 1987 6 2.76 0.36 2.38–3.14 6 2.68 0.36 2.30–3.06 or Laetoli UNB this work 7 2.43 0.15 1.86–3.10 7 2.27 0.18 1.71–3.10 DP/4 LB Xerus sp. this work 2 2.14 2.04–2.24 2 2.15 1.91–2.38 Olduvai 6 3.13 0.42 2.30–3.40 6 3.07 0.51 2.05–3.45 Omo B, C 2 2.50–2.60 2 2.76–2.88 KB-97-162 1 2.41 1 2.31 M/1 Laetoli UNB 1987 8 3.13 0.16 2.99–3.26 8 3.21 0.3 2.96–3.46 Laetoli UNB this work 9 2.93 0.07 2.62–3.3 9 2.87 0.08 2.52–3.19 LB Xerus sp. this work 3 2.46 2.38–2.5 3 2.56 2.5–2.62 Olduvai 9 3.48 0.14 3.30–3.65 9 3.51 0.19 3.30–3.80 Omo B, C a 2 3.12–3.46 2 3.60–3.70 KB-97-162 1 2.62 1 2.9 Kanapoi 2 2.54 2.54–2;55 2 2.85 2.84–2.86 M/2 Laetoli UNB 1987 3 3.32 0.16 2.92–3.72 3 3.47 0.12 3.47–3.77 Laetoli UNB this work 7 3.03 0.07 2.76–3.29 7 2.99 0.09 2.62–3.3 LB Xerus sp. this work 2 2.52 2.38–2.65 2 2.66 2 2.60–2.71 Olduvai 6 3.63 0.15 3.45–3.80 6 3.57 0.13 3.57–3.7 KB-97-162 1 2.76 1 3.17 Kanapoi 2 2.57 2 2.98 2.94–3.01 M/3 Laetoli UNB1987 2 3.78 2 3.3 Laet. UNB this work 4 3.1 0.1 2.86–3.33 4 2.98 0.06 2.81–3.1 Olduvai 3 3.55 0.09 3.45–3.6 3 3.5 0.17 3.4–3.7 Omo F 2 3.25–3.5 2 2.6–2.77 KB-97-162 1 2.97 1 2.66 Kanapoi 2 2.49 2.49–2.5 2 2.78 2.76–2.79 Data sources: Olduvai Bed I, X. cf. inauris, Denys (1990); Omo B, C, Xerus erythropus, Wesselman (1984); Omo F, Xerus sp., Wesselman (1984); KB-97-162, Chad, Xerus daamsi, Denys et al. (2003); Kanapoi, Xerus sp., Manthi (2006). Abbreviations: UNB, Upper Ndolanya Beds; LB, Laetolil Beds a Attribution to M1 or M2 is ambiguous. DP/4 and P/4 have been pooled, which may explain the high variability observed for these teeth
Xerus sp. (Fig. 2.4, Table 2.5) The Upper Laetolil Beds have also yielded the remains of a smaller sciurid. This species is very bunodont and was described and figured by Denys (1987a) from Loc. 9S (LAET 75-1562, Plate 6.2) and possibly includes the
Gadjingero 100 skull (from the Kohl-Larsen collection in Berlin). New specimens from Loc. 9 (EP 1089/98) and Loc. 9S (EP 1215/04) can be attributed to this same species. The entoconid and posterolophid are very crestiform on P/4, M/1, and M/2, and they make a continuous distal wall on M/2. There is no prominent entoconid and the cusp relief is low. These features are similar to modern
24
C. Denys
X. rutilus (Denys et al. 2003). The specimens display a shorter skull length, relatively longer upper molar row (Fig. 2.4) and smaller lower molars than X. janenschi (Tables 2.4 and 2.5), although measurements of some teeth fall in the low end of the range of X. janenschi. Family Nesomyidae Major, 1897 Subfamily Cricetomyinae Roberts, 1951 Tribe Saccostomurini Roberts, 1951 Genus Saccostomus Peters, 1846 Saccostomus major Denys, 1987 (Figs. 2.7–2.12) Many new specimens from the Upper Laetolil Beds (Locs. 1, 2, 3, 4, 5, 6, 7, 8, 9S, 10, 10E, 10W, 11, 15, 17 and 22) are attributed to S. major (see Appendix 2.3). They display the same morphological characteristics of the teeth as the previously recovered material (Fig. 2.7). The initial description of the species included nearly complete skulls, and there are no new skeletal elements to describe here. However, with the recovery of 173 new Saccostomus individuals from the Upper Laetolil Beds we have been able to study the population at a finer scale. M/1 length and wear stages were analyzed to assess the variability among the species and to detect biostratigraphic differences. The following wear stages can be defined (Fig. 2.8):
Fig. 2.8 Wear stages of the M1/ and M/1 Saccostomus major from the Upper Laetolil Beds. Top row, M1/. (a) stage 1, EP 1375/00; (b) stage 2, EP 160/03; (c) stage 3, EP 3904/00; (d) stage 4, EP 998/05. Bottom row, M/1. (e) stage 1, EP 162/03; (f) stage 2, EP 1424/03; (g) stage 3, EP 2434/03; (h) stage 4, EP 1065/03. Scale bar indicates 1 mm
Stages 0–1: Presence of two isolated cusps on the prelobe of M1/1 or M3/ unerupted. Stages 2–4: Cusps visible on all the molars, no large longitudinal links between cusps visible. Stages 5–6: Wide links between the lobes and cusps hardly visible on the M3/ and the whole tooth row. Among the newly collected material of S. major one finds a good proportion of juveniles (stages 0–1: 33.3%) and old adults (stages 5–6: 21.6%) compared to prime adults (stages 2–4: 45.1%).
Fig. 2.9 Scatterplot of the M/1 dimensions (mm) of Saccostomus major by locality. The specimens from Loc. 7E (black triangles) come from the Upper Ndolanya Beds
Fig. 2.7 Left maxillary toothrow of Saccostomus major from the Upper Laetolil Beds. EP 1738/04 from Loc. 2 (left) and EP 1326/03 (right M1/) from Loc. 11 (right). Scale bar indicates 1 mm
The scatter plot of M/1 length by width, organized by locality, does not provide a clear pattern of size differences (Fig. 2.9). Specimens from the pooled Loc. 10 complex of localities encompass the full range of variation, while it appears that specimens from Locs. 1, 3, 5, 9, and 11 are slightly larger than those from Locs. 6, 8 and 15. When the data are sorted by stratigraphic level (i.e., below Tuff 2, below Tuff 3, between Tuffs 3–5, between Tuffs 5–7, between Tuff 7 and the Yellow Marker Tuff) one observes a slight
2 Rodents
decrease in size between the lower levels and upper levels (Fig. 2.10). However, the sample is too small to reach a definitive conclusion about the biostratigraphic variation of Saccostomus M/1 through the Upper Laetolil Beds. Specimens from the Upper Ndolanya Beds fall in the middle of the distribution. We confirm the presence of S. major in the Upper Laetolil Beds and add it to the faunal list of Loc. 15. However, it is still absent from Locs. 12, 13 and 21, as Denys (1987a) previously observed. It is not yet found in the Lower Laetolil Beds, although it is recorded at older eastern and southern African sites. Saccostomus major is described from the Manonga Valley (Winkler 1997), while S. geraadsi was named by Mein et al. (2004) from Ch’orora (Ethiopia) and Harasib 3a (Namibia). Finally, Mein and Pickford (2006) recognized S. cf. geraadsi based on molars from Lukeino in Kenya, dated to around 6.1–5.8 Ma. Saccostomus major from Laetoli is similar in size to that from the Manonga Valley (Table 2.6, Fig. 2.11). Saccostomus geraadsi from Lukeino and Harasib have smaller molars compared to S. major, and their molar size fits within the variability of S. cf. mearnsi from Olduvai Bed I (Fig. 2.11). Saccostomus cf. major (Figs. 2.9–2.12) Denys (1987a) described a single mandibular fragment (LAET 75-862) from Loc. 18 (Upper Ndolanya Beds) and left it unattributed at the species level due to the small size of the M/2-3. New remains of Saccostomus have been recovered from the Upper Ndolanya Beds at Loc. 7E and are described here.
25
Referred material: Loc. 7E. EP 1247/03 (Fig. 2.12), isolated right M/1. EP 1248/03, associated mandibles. EP 1249/03, left mandible fragment with M/1-2. The isolated lower molar belongs to a young individual (wear stage 2) with the two cusps of the prelobe still visible (Fig. 2.12). It is comparable in size to specimens from the Upper Laetolil Beds (Fig. 2.9). Because it shows dentine and enamel corrosion we cannot describe the specimen in detail, except to mention that it has a link between the prelobe and the first lobe and a tiny cingular cV5 on the labial side of the molar. The mandibular fragment with M/1-2 also fits within the size variation of other Laetoli S. major specimens, and can be attributed to wear stage 1 (Fig. 2.9). The main differences distinguishing the Laetoli material from S. cf. mearnsi of Olduvai are the large prelobe of M/1 and the presence of an anterolabial crest (absent in the Olduvai Bed I specimens). Consequently, the new Saccostomus specimens from Loc. 7E can be attributed to S. cf. major pending additional finds from the Upper Ndolanya Beds. Because no new material was recovered from Loc. 18 we retain here Saccostomus sp. for the unique specimen from the Mary Leakey collection. Saccostomus cf. major A single mandibular fragment (EP 2075/03) with a broken M/1 (with trace of two roots) and a well-preserved M/2 is known from Emboremony 1 (Lower Laetolil Beds). This molar is of wear stage 4 and displays two relatively transverse lobes with fused cusps and an anterolabial cingulum. Its size (1.72 × 1.81 mm) falls within the range of the M/2s of S. major from the Upper Laetolil Beds. Family Muridae Illiger, 1811 Subfamily Gerbillinae Gray, 1825 The Upper Laetolil Beds have already yielded two different species of Gerbillinae (Denys 1987a). One (Gerbillinae sp.) was not attributed to any genus due to the low number of specimens and the limited availability of characters. The other was attributed to Gerbilliscus cf. inclusa and was characterized by wide molars, very transversely aligned cusps, and mesially open prelobe on M/1. The new collections allow a more detailed description of the Gerbillinae sp. of Denys (1987a). Recent molecular revisions have changed the genus nomenclature, so we follow Wilson and Reeder (2005) in retaining Gerbilliscus for the Laetoli specimens in place of the old name Tatera. Genus Gerbilliscus Thomas, 1897 Gerbilliscus satimani sp. nov. (Figs. 2.13–2.16)
Fig. 2.10 Scatter plot of Saccostomus major M/1 grouped by stratigraphic level: ULB, Upper Laetolil Beds; 3-5 = between Tuffs 3 and 5; 6-7 = between Tuffs 6 and 7; 7-8 = between Tuffs 7 and 8; 7-YMT = between Tuff 7 and Yellow Marker Tuff; Lower 2 = below marker tuff 2; Lower 3 = below Tuff 3, UNB, Upper Ndolanya Beds
Holotype: EP 147/01, nearly complete cranium with associated mandibles. Nasal region missing (Fig. 2.13). Type locality: Loc. 6, Laetoli, Tanzania. Age and horizon: Mid-Pliocene, Upper Laetolil Beds.
Laetoli this work Laetoli (Denys 1987a) Olduvai Bed I Manonga Harasib 3 Lukeino
Laetoli this work Laetoli (Denys 1987a) Olduvai Bed I Harasib 3
Laetoli this work Laetoli (Denys 1987a) Olduvai Bed I Manonga Harasib 3 Lukeino
Laetoli this work Laetoli (Denys 1987a) Olduvai Bed I Manonga Harasib 3
M2/
M3/
M/1
M/2
20 34 135 2 41
56 38 139 3 42 7
4 5 20
0.02
0.01
1.67
0.01 0.04
0.02
0.03 0.04 0.02
0.01
0.04
0.02
0.03
SD
1.86 2.06 1.69
2.78 2.78 2.35 2.71 2.22 2.23
1.27 1.13 1.11
1.86 2.01 1.63 1.96 1.68
2.51
40 3 8 13 61 3 44 2
3.00 3.00 2.66
Mean
16 21 43
N
1.67–2.00 1.85–2.33 1.54–1.89 1.84–1.88 1.49–1.85
2.40–3.10 2.48–3.03 2.19–2.57 2.60–2.83 2.05–2.40 2.09–2.35
1.22–1.35 1.00–1.22 0.93–1.29
1.68–2.00 1.80–2.37 1.53–1.87 1.88–2.04 1.44–1.83 1.56–1.60
2.30–2.79 2.24–2.40
2.86–3.19 2.83–3.38 2.52–2.81
Range
20 34 134 2 41
56 38 140 5 42 7
4 5 20
8 13 61 3 44 2
16 21 46 3 40 3
N
1.56
1 0.83 1.93 1.65
1.81 1.83 1.57 1.83 1.51 1.46
1.48 1.24 1.16
1.81 1.9 1.67 2.05 1.65
1.93 2.08 1.77 2.19 1.68
Mean
0.01
0.02
0.01 0.02
0.01
0.13 0.02 0.02
0.01
0.03
0 .01
0.03
SD
1.57–2.00 1.75–2.15 1.49–1.78 1.84–1.96 1.43–1.70
1.55–2.00 1.63–2.00 1.42–1.77 1.75–1.92 1.36–1.69 1.41–1.54
1.30–1.85 1.18–1.29 0.93–1.28
1.71–2.00 1.78–2.07 1.57–1.84 2.00–2.12 1.47–1.83 1.58
1.67–2.14 1.80–2.40 1.63–1.91 2.16–2.25 1.53–1.81 1.57–1.65
Range
Laetoli this work 12 1.48 0.04 1.33–1.76 12 1.39 0.03 1.19–1.57 Laetoli (Denys 1987a) 4 1.64 0.04 1.55–1.72 4 1.59 1.48–1.80 Olduvai Bed I 17 1.28 0.02 1.16–1.41 17 1.24 0.02 1.16–1.37 Lukeino 1 1.26 1 1.32 Harasib 3 30 1.3 0.02 1.18–1.46 30 1.19 0.01 1.08–1.39 N number of molars, SD standard deviation Data sources: previous study of Laetoli (Denys 1987a), S. cf. mearnsi, Olduvai Bed I (Denys 1992), S. major, Manonga (Winkler 1997), S. geraadsi, Harasib 3 (Mein et al. 2004), S. cf. geraadsi, Lukeino (Mein and Pickford 2006)
Laetoli this work Laetoli (Denys 1987a) Olduvai Bed I Manonga Harasib 3 Lukeino
M1/
M/3
Locality
Tooth
Table 2.6 Molar dimensions (mm) of Saccostomus spp. from Laetoli and from other Plio-Pleistocene sites Length Breadth
26 C. Denys
2 Rodents
27
Saccostomus upper M1/ 2,4 2,3
S. major Laetoli
2,2
Width
2,1
S. cf. mearnsi Olduvai
2 1,9
Harasib S. geraadsi
1,8 1,7
Lukeino S. geraadsi
1,6 1,5
2
2,2
2,4
2,6
2,8
3
3,2
3,4
Length Saccostomus Lower M/1 2,1 2
Olduvai
Width
1,9
Harasib
1,8
Manonga Lukeino
1,7
Laetoli
1,6 1,5 1,4 1,8
2
2,2
2,4
2,6
2,8
3
3,2
3,4
Length Fig. 2.11 Comparison of fossil Saccostomus spp. M1/ and M/1 from different localities. Length and width in mm. After Winkler (1997), S. major Ibole, Manonga Valley; Mein et al. (2004), S. geraadsi, Harasib;
Mein and Pickford (2006), S. geraadsi, Lukeino; Denys (1987a, this study) S. major, Laetoli; Denys (1992), S. cf. mearnsi, Olduvai Bed I
Type series: EP 782/03, left mandible with M/1-2, Loc. 9. EP 999/01, left maxilla with M1/, Loc. 11. EP 1075/04, left mandible with M/1 (Fig. 2.16), Loc. 1. EP 1981/03, anterior cranial fragment with right maxilla with M1-2/, Loc. 7 (Figs. 2.14 and 2.16). EP 1889/03, left mandible with M/1-3, Loc. 1 (Fig. 2.14). Referred material: LAET 75/A17, M1/, Loc. 6. LAET 75-3492, M1/, Loc. 10W (Plate 6.2 in Denys 1987a). LAET 79/A02, M1/, Loc. 6 (Plate 6.2 in Denys 1987a). LAET 79/A3761, M/1, Loc. 6. LAET 79/A13B, M/1, Loc. 6. LAET 79/A5B1, M/1, Loc. 5. LAET 79/A5B2, M/1, Loc. 5. LAET 79/A13, mandibular fragment with M/1-2, Loc. 6. Distribution: Locs. 1, 5, 6, 7, 9 and 11 of the Upper Laetolil Beds.
Measurements: Table 2.7 and 2.8, Fig. 2.16 Repository: National Museum of Tanzania, Dar es Salaam Etymology: Named after Satiman, the volcano probably responsible for producing the volcanic ash at Laetoli that allowed such exceptional preservation. Diagnosis: A Gerbilliscus with quite narrow molars, cusps distinguishable, simple rounded prelobe of M1/1 open anteriorly on M/1 when unworn. No longitudinal link between the prelobe and first lobe of the M/1. Long palatal foramen (from the first lobe of M1/ to the front of the second lobe of M2/). No posterior cingulum visible on M/1. Small bilobed M3/. Well-developed tympanic bullae. Differs from G. gentryi from Olduvai Bed I by its much smaller size (Fig. 2.15), the anterior opening of the M/1
28
p relobe, and well-individualized cusps. Differs from Gerbillus spp. from Olduvai Bed I and from Late Miocene site of Asakoma (Middle Awash, Ethiopia) by the absence of a longitudinal link between the prelobe and the first lobe of the M/1 and by the transversely aligned cusps. Differs from Gerbillus sp. from Omo Members B and F by a more rounded and larger prelobe of M1/, less fused cusps and by the anterior opening of the prelobe. Differs from Gerbillus sp. of Lemudong’o and Kanapoi by its much larger-sized molars.
Fig. 2.12 Saccostomus cf. major from the Upper Ndolanya Beds (Loc. 7E), EP 1247/03. Scale bar in mm
C. Denys
Description: The holotype consists of a well-preserved cranium, but the rostrum is broken and the upper incisors are absent (Fig. 2.13). Another specimen (EP 1981/03, Loc. 7) displays the premaxilla, maxillary toothrows and nasal bones with in situ incisors that show a median groove. The interorbital constriction is poorly marked. The right tympanic bulla is nearly complete. The latter is inflated in both tympanic and mastoid regions, as in modern Gerbilliscus (ex Tatera) and Gerbillus. The incisive foramen is short and stops far from the anterior root of the M1/, while the palatal foramen, which is long, begins at the level of the first lobe of the M1/ and ends at the front of the second lobe of M2/ (Fig. 2.13). The holotype is an old individual and its molars are quite worn, but the cusps are still visible. The prelobe of M/1 is round or composed of two cusps separated by a deep anterior groove (the so-called anterior opening). The M1/ displays a round and narrow prelobe, with a distal crest not related to the first lobe. There is also the trace of a distal crest on the first lobe of M1/ with the two cusps not well aligned in a transverse lamina (Figs. 2.14 and 2.16). There is no distal cingulum on the type specimen and only one specimen displays a trace. Similarly, there is no anterocone on the M2/. The M3/ is composed of two lobes of nearly equal size and the crown is not very reduced in overall size. Gerbilliscus satimani sp. nov. is slightly smaller than extant G. leucogaster from South Africa and has narrower molars (Table 2.7). The two species share the prelobe anterior opening on M/1. The disposition of the incisive and palatal foramina is similar. The cranial proportions are comparable between the two species for molar row length and interorbital constriction, but the tympanic bullae of G. satimani sp. nov are more developed than in G. leucogaster and
Fig. 2.13 Dorsal (a) and ventral (b) view of the holotype of G. satimani nov. sp. Cranium EP 147/01 from Loc. 6, Laetolil Beds. Scale bar 1cm
2 Rodents
29
Fig. 2.14 SEM images of G. satimani specimens. Left, EP 1981/03 from Loc. 7 (left M1-2/). Right, EP 1889/03 from Loc. 1 (left mandible with M/1-3). Scale bar in mm
equivalent in size to those of the extant South African Gerbillurus vallinus (Table 2.7). However, the modern South African Gerbillurus has a much shorter molar row than G. satimani (Table 2.7). Gerbilliscus satimani is smaller than G. cf. inclusus from the Upper Laetolil Beds and the new Gerbilliscus species from the Upper Ndolanya Beds (Figs. 2.15 and 2.16). Gerbilliscus satimani differs from G. gentryi from Olduvai bed I in being slightly smaller in size, and having narrower molars, a prelobe opening anteriorly, and the retention of a slight trace of a longitudinal crest on the upper molars (Figs. 2.15 and 2.16). Gerbilliscus satimani also differs from Tatera sp. (= Gerbilliscus sp.) from the late Miocene of Asakoma, Middle Awash (Ethiopia) by its larger molars, the absence of a longitudinal crest, wellindividualized transverse cusps, and a prelobe with two unfused cusps of unequal size (Wesselman et al. 2009). Mein and Pickford (2006) described a new species of Abudhabia from Kapsomin in the Lukeino Formation, mentioning that it may be intermediate between Abudhabia and Tatera sensu stricto (= Gerbilliscus) and similar in size and morphology to the Gerbillinae indet. of Laetoli of Denys (1987a). By comparing the new specimens to the figured one, we find that the size is similar, but there are a lot of morphological differences, which prevent one from recognizing a close affinity between the two species. Among these differences are the absence of anterocone and anteroconid on M2/2 and the quasi absence of a posterior cingulum on M/1, the absence of a groove between the two cusps on the M1/ prelobe and their fusion with a rounded aspect (which is a Gerbilliscus character). The main difference concerns the prelobe of M/1, which displays an anterior opening on poorly worn specimens of G. satimani
or simply a rounded prelobe that is very different from the specimens figured by Mein and Pickford (2006). The Laetoli specimens also do not fit well with the Gerbilliscus sp. material from Hadar described by Sabatier (1982) because of the M/1 prelobe opening, which is located posteriorly in the latter specimens. The Hadar specimens also retain a trace of a posterior cingulum on M1/ and a small anteroconulid on M/2, as well as distinct cusps. Modern representatives of Gerbilliscus (G. leucogaster and G. nigricauda) may display traces of a posterior cingulum on M1/, so this cannot be taken as a valid character to distinguish Abudhabia from Gerbilliscus (= Tatera). Wesselman (1984) described Tatera sp. indet. (= Gerbilliscus) from Omo Members B and F. The specimens from Omo Member B share with G. satimani the relatively well-individualized cusps of the first lobe of the M1/, but the former have a prelobe on the M1/ with an anterior depression in the middle, and traces of the two cusps that constitute it. The size of the molars is similar to those of G. satimani, but the M3/ is bilobed, whereas it is small in the Laetoli fossils (Table 2.8). The lower molars from Omo Member B are also like those figured from Omo Member F, and they display a different shaped prelobe on M/1 (posterior opening) and are nearly equal in size to G. gentryi specimens from Olduvai Bed I. Manthi (2007) described a Gerbilliscus (Tatera) sp. from Lemudong’o and the figured specimens display worn molars. The size of M/1 (Table 2.8) is much smaller than those of G. satimani. The M1/ of Gerbilliscus sp. from Kanapoi described by Manthi (2006) displays a round prelobe and a trace of cusps on the first row. They are also small, being similar in size to those from Lemudong’o (Table 2.8). These specimens may fit within the G. satimani
30
C. Denys
Table 2.7 Skull measurements for modern and fossil Gerbilliscus and Gerbillurus species A G. satimani EP 147/01 EP 1889/03 G. cf. inclusus EP 1372/98 LAET 75-3588 G. gentryi Olduvai Bed I
G. leucogaster Tanzania & South Africa
G. vallinus Namibia & SW Africa NHM95-331 NHM25-1-2-87 NHM25-1-2-85
5.11 5.51
B 5.28
C 6.33
D 11.7
7.2 7.2 N Min Max Mean SD
13 5.30 5.91 5.61 0.21
1
N Min Max Mean SD
48 4.94 6.38 5.55 0.32
46 5.28 6.54 5.91 0.24
3.83
11.89 11.75 11.63
10.5
5.4
4.14
4.12 4.02 4.28
G. afra Angola NHM29-10-1-19
5.2
6.1
G. paeba South Africa NHM3-1-4-27 NHM49-345
4.04 3.97
4.05 4.22
48 5.64 6.94 6.10 0.32
48 9.07 11.20 10.19 0.57
Fig. 2.15 Size comparisons of M/1 between Olduvai Bed I Gerbilliscus gentryi (Levels K, L2, L3, M1, M2, M3, M4, M5, M6 after Denys 1989a), Gerbilliscus spp. from Hadar (A.L. 333 and A.L. 327, mean value after Sabatier 1982) and Asakoma (ASK-MA after Wesselman et al. 2009), with G. cf. inclusus, G. satimani sp. nov., G. winkleri sp. nov. of Laetoli
8.32 9.39
G. swalius SW Africa NHM25-12-4-110 3.86 4.12 7.92 Data from the literature for Olduvai Bed I (Denys 1989a) and modern specimens from museum collections A LI13, length of the lower molar row B LS13, length of the upper molar row C CIO, interorbital constriction width (taken in dorsal view) D LBT, length of the tympanic bulla
lineage or belong together in a new smaller species as yet undescribed.
Fig. 2.16 Comparison of the upper and lower M1 of the different species of Gerbilliscus from Laetoli. Upper row, M/1. (a) G. satimani, EP 1075/04; (b) G. winkleri, EP 3319/00; (c) G. winkleri, EP 3500/00, (d) G. cf. inclusus, EP 1372/98. Lower row, M1/. (e and f) G. satimani EP 1981/03 and EP 999/01; (g) G. winkleri, EP 3320/00. Drawn to the same scale; the scale bar indicates 1 mm
Gerbilliscus winkleri nov. sp. (Figs. 2.15–2.17) Additional specimens from the Upper Ndolanya Beds (Locs. 18 and 15) allow attribution of the previously so-called Gerbilliscus (Tatera) sp. from Laetoli and Hadar to a newly recognized species. Holotype: EP 3320/00, left mandible with M/1 and associated right maxilla with M1-2/ (Fig. 2.16) Type locality: Loc. 18, Laetoli, Tanzania. Age and horizon: 2.66 Ma, mid-Pliocene, Upper Ndolanya Beds.
Etymology: in honor of Alisa Winkler who has described numerous fossil rodents from East African Neogene sites. Type series: Loc. 18, Upper Ndolanya Beds: EP 3319/00, left mandible with M/1-2. EP 817/01, left mandible with M/1 (Fig. 2.17). EP 3520/00, left maxillary fragment with M1-2/ and lower left M/1. Loc. 15, Upper Ndolanya Beds: EP 3500/00, left mandible with M/1. Referred material: Previous Tatera sp. collections from Locs. 7E and 18, Upper Ndolanya Beds (after Denys 1987a):
Locality
G. satimani this work G. satimani Denys, 1987a G. winkleri this work G. winkleri Denys, 1987a G. gentryi Denys, 1989a G. sp. Lemudong’o G. sp. Hadar G. sp. Omo B G. sp. Omo F G. sp. ASK-VP3 G. sp. Kanapoi
G. satimani this work G. winkleri this work G. winkleri Denys, 1987a G. gentryi Denys, 1989a G. sp. Lemudong’o G. sp. Hadar G. sp. Omo F
G. satimani this work G. gentryi Denys, 1989a G. sp. Hadar G. sp. Omo F
G. satimani this work G. santimani Denys, 1987a G. winkleri this work G. winkleri Denys, 1987a G . cf. inclusus this work G. cf. inclusus Denys, 1987a G. sp. Lemudong’o G. gentryi Denys, 1989a G. sp. Hadar G. sp. Omo B G. sp. Omo F G. sp. AMW-VP1 G. sp. Kanapoi
Tooth
M1/
M2/
M3/
M/1 5 4 3 8 1 2 7 28 19 1 2 1 6
1 1 1 1
1 1 1 7 1 16 1
2 3 1 2 23 1 19 4 3 1 4
N
Length
Table 2.8 Molar dimensions of fossil Gerbilliscus spp.
2.54 2.77 2.95 3.29 3.86 3.77 1.79 2.90 3.17 2.72 2.79 3.20 2.05
0.86 0.85 1.08 1.06
1.52 1.86 1.85 1.69 1.00 1.85 1.82
0.14
2.94 1.90 3.20 2.60 2.73 2.95 2.23
3.43–4.1 1.6–2.00 2.76–3.11 3.01–3.38
0.47 0.06 0.1 0.09
1.9–2.20
2.67–2.90
2.43–2.72 2.65–2.88 2.88–3.05 3.1–3.62
1.66–1.98
0.09
0.11 0.12 0.08 0.16
1.57–1.81
2.10–2.52
3.01–3.40 2.32–2.77 2.52–2.95
3.12–3.18 2.62–3.18
2.76–2.81 2.55–2.62
Range
0.09
0.09
0.06
SD
2.62 2.95
Mean
5 4 3 9 1 1 7 28 19 1 4 1 6
1 1 1 1
1 1 1 7 1 16 1
2 3 1 2 23 1 19 4 1 1 4
N
Breadth
1.64 1.84 1.92 2.16 2.57 2.70 1.30 1.87 1.98 1.64 1.72 2.12 1.48
1.10 1.29 1.40 1.27
1.67 2.05 2.05 1.90 1.20 1.98 1.90
2.15 1 2.16 1.88 2.07 2.17 1.65
1.98 2.15
Mean
0.03 0.09 0.06
0.11 0.07 0.07 0.07
0.06
0.11
0.06
0.08
0.03
SD
1.40–1.50 (continued)
1.68–1.80
1.20–1.40 1.69–2.05 1.89–2.06
1.52–1.81 1.80–1.95 1.86–2.00 2.02–2.28
1.89–2.08
1.74–2.03
1.6–1.81
2.03–2.25 2.79–1.98
2.33–2.45 1.96–2.28
1.67–1.86 1.95–2.00
Range
2 Rodents 31
4 1 1 5 3 1 10 19 1 3
N
Length 1.50 1.55 1.71 2.05 1.07 2.12 1.74 1.85 1.87 1.27
Mean
1.9–2.35 1.00–1.20 1.57–1.87 1.75–1.96
0.17 0.07 0.09 0.05 1.20–1.40
1.43–1.57
Range
0.06
SD 4 1 1 5 3 1 10 19 1 3
N
Breadth 1.61 1.70 1.95 2.11 1.30 2.52 1.90 1.92 1.90 1.27
Mean
0.08 0.07
0.04 0.06
0.11
SD
1.20–1.40
1.76–2.00 1.79–2.05
2.05–2.15 1.20–1.40
1.52–1.76
Range
G. satimani this work 3 0.56 0.4 0.57–0.91 3 1.11 0.12 1.00–1.24 G. gentryi Denys, 1989a 3 1.06 0.1 0.96–1.15 3 1.25 0.09 1.16–1.33 G. sp. Hadar 2 0.96 0.93–1.00 2 1.29 1.27–1.32 G. sp. Omo F 1 1.10 – G. cf. inclusus 1 1.20 1 1.78 N number of molars, SD standard deviation Data sources: Hadar (Sabatier 1982), Omo Shungura B, F, G (Wesselman 1984), Lemudong’o (Manthi 2007), Middle Awash (Asa Koma) (Wesselman et al. 2009), Kanapoi (Manthi 2006)
M/3
Locality
G. satimani this work G. satimani Denys, 1987a G. winkleri this work G. winkleri Denys,1987a G. sp. Lemudong’o G. cf. inclusus G. gentryi Denys, 1989a G. sp. Hadar G. sp. Omo G G. sp. Kanapoi
Tooth
M/2
Table 2.8 (continued)
32 C. Denys
2 Rodents
LAET 75-728 (Plate 6.2 in Denys 1987a), LAET 75-899, LAET 75-673, LAET 76/71-72, LAET 75-862, LAET 75-602, LAET 74-36, LAET 75-636, LAET 74-35, LAET 75-661, and LAET 75-894 (Plate 6.2 in Denys 1987a). Diagnosis: Well-aligned transverse cusps poorly individualized (fused into transverse laminae), prelobe of M/1 rounded or open distally. Oval-shaped prelobe of M1/, with no trace of cusps. Larger molars than G. gentryi and G. satimani. Same size as modern G. leucogaster from South Africa, but with much more fused and transverse laminae and a longer palatal foramen. In G. winkleri the palatal foramen starts at the level of the second lobe of the M1/ and ends at the back of the M2/, while in G. leucogaster the palatal foramen is situated between the first and second lobes of the M2/. Differs from G. gentryi Denys, 1990 from Olduvai Bed I and Gerbilliscus sp. from Omo Members B and F in the larger size of M1/1. Differs from Gerbilliscus sp. from Hadar in the smaller M1/. Differs from Gerbilliscus sp. from Asakoma (Middle Awash, Ethiopia) by the prelobe of the M/1 displaying two cusps of unequal size and more fused cusps. Differs from G. satimani sp. nov. in the larger size of the molars and M/1 prelobe with well fused cusps and distal opening of the M/1 prelobe.
33
Fig. 2.17 Gerbilliscus. winkleri sp. nov. EP 817/01 (paratype) from Loc.18 (Upper Ndolanya Beds), left mandibular fragment with M/1. Scale bar indicates 1 mm
Measurements: Table 2.8, Fig. 2.15. Description: In the Upper Ndolanya Beds one finds a somewhat larger Gerbilliscus, which has more transversely aligned cusps than in G. satimani. It has quite large molars with generally well-fused cusps in transverse laminae, especially the first lobe of M1/1 (Fig. 2.17). Either the prelobe of the M/1 is rounded (55% of cases) or, on one unique unworn specimen (EP 3319/00), it is composed of two cusps of equal size, linked anteriorly and separated by a deep posterior groove, giving a horseshoe-shaped configuration (27%) (Figs. 2.16 and 2.17). Only 9% of specimens have a prelobe open distally, compared with 26% in G. gentryi from Olduvai. No M3/3s are yet known for this species. The M1/1s are larger on average than those of G. gentryi and Gerbilliscus sp. from Omo Members B and F, with which they may be related (Table 2.8, Fig. 2.15). The Hadar specimens (A.L. 333 and A.L. 327 localities) are very similar to the Laetoli G. winkleri in the shape of the M/1 prelobe, and fall just at the size limit between G. gentryi and G. winkleri (Fig. 2.15). We only used the average value provided in Sabatier (1982). Further knowledge of the range of variability of the A.L. 327 sample should help resolve whether or not the Hadar specimen can be placed in synonymy with G. winkleri. The Middle Awash ASK-VP1 unique M/1 fits within the range of variability of G. winkleri (Fig. 2.15). It belongs to an unworn molar and in contrast to G. winkleri specimens it exhibits the trace of two unequal size cusps (unfused) on the M/1, which prevents synonymy with either the Laetoli or Hadar taxa.
Fig. 2.18 Gerbilliscus cf. inclusus (EP 1372/98) from Loc. 13. Mandible fragment with M/1-3. Scale bar indicates 1 mm
Gerbilliscus cf. inclusus (Figs. 2.15–2.18) One new mandible fragment with M/1-3 (EP 1372/98), distinguished by its larger size (Fig. 2.15), has been recovered from Loc. 13 in the Upper Laetolil Beds (Fig. 2.18). Its large dimensions fit with those of Denys (1987a) specimens LAET 75-2726 (Loc. 3) and LAET 75-3588 (Loc. 8), and we add one new taxon to the faunal list of Loc. 13 (Fig. 2.15, Table 2.8). This new specimen has an M/1-3 length of 7.2 mm, which is the same as that for LAET 75-3588 (Table 2.7). The prelobe of M/1, which is a key character for species identification, was not well
34
C. Denys
preserved on specimens from the old Laetoli material, but on the new specimen it is rounded in its anterior part and distally elongated. The M/3 has only one lobe and is very reduced. Cusps are well fused and transverse, and there is no trace of cingular cusplets, but on the M/2 one still recognizes a trace of a longitudinal link between first and second lobe. The prelobe of M/1 of the Upper Laetolil Gerbilliscus cf. inclusus is different from that in Gerbilliscus winkleri sp. nov. from Laetoli, G. gentryi from Olduvai Bed I or Gerbilliscus sp. of Asakoma site (Middle Awash). Denys (1987a) compared these specimens with various modern Gerbilliscus representatives and found similarities with G. inclusus, due to the disposition of the prelobe of the M/1 (open anteriorly) and the relatively small proportions of M/3. Among Gerbilliscus of large size (afra group of Meester et al. 1986) one also finds G. afra and G. brantsi, which display a prelobe open anteriorly and with well-aligned cusps. Gerbilliscus brantsi displays a wider M/1 compared to G. inclusus and G. afra, but molar variability is not well known, so pending further taxonomic revisions of this complex we prefer to keep these rare fossils at Laetoli unassigned and retain the initial attribution of Denys (1987a).
2. EP 654/03, right mandible with M/1, Loc. 2. EP 1783/03, right mandible with M/1, Loc. 22. EP 996/05, left mandible with M/1-2, Loc. 2. EP 2239/00, left mandible with M/1-3, Loc. 7. EP 1739/04, right mandible with M/1 and a broken M/2, Loc. 2. EP 1871/03, left mandible with M/3, Loc. 4. EP 243/05, left mandible with M/1, Loc. 9. This small stephanodont murid is quite easy to identify with its stephanodont molars, its small size, and the presence of accessory roots on the M/1. It is rather abundant in the new collections and is found, as in the previous collections, at Locs. 5, 6 and 9. It is also found for the first time at Locs. 2, 4, and 22, but this time was not recorded at Locs. 10, 11 and 21. The new specimens comprise only lower molars and they fit well with the dimensions of the previously collected material (Table 2.9). There is a large size variation within this species that is not explained by locality
Subfamily Murinae Illiger, 1811 Thallomys laetolilensis Denys, 1987 (Fig. 2.19) Localities and horizons: Locs. 2, 4, 5, 6 and 9. Upper Laetolil Beds up to Tuff 7. Referred material: EP 148/01, left mandible with M/1-3, Loc. 6. EP 2034/03, right mandible with M/2, Loc. 6. EP 2033/03, left mandible with M/1-3, Loc. 6 (Fig. 2.19). EP 1039/05, right mandible with M/2, Loc. 2. EP 244/05, right mandible with M/1-2 (very worn), Loc. 9. EP 243/05, right mandible with M/1-2, Loc. 9. EP 397/03, right mandible with M/1-3, Loc. 5 (Fig. 2.19). EP 1065/03, right mandible with M/1-3 (very worn), Loc. 10W. EP 187/03, right mandible with M/3, Loc. 4. EP 655/03, right mandible with M/1, Loc.
Fig. 2.19 Thallomys laetolilensis. Left, EP 397/03 from Loc. 5; right, EP 2033/03 from Loc. 6. Scale bars indicates 1 mm
Table 2.9 Length of M/1-3 and length and width dimensions of M/1 (mm) of the new specimens of Thallomys laetolilensis compared to the mean values of the type series (Denys 1987a) Material
Specimen
LM/1-3
New specimens
EP 148/01 EP 397/03 EP 2033/03
5.32 4.84 4.82
M/1L
M/1W
Previous Laetoli collections (type series) (Denys 1987a) (N = 11)
Mean Range
New specimens (N = 7)
Mean Range
2.02
1.24
(1.91–2.19)
(1.14–1.38)
SD
0.04
0.03
2.15 (2.00–2.30) 0.12
1.34 (1.20–1.47) 0.08
Previous Laetoli collections (type series) (Denys 1987a, Table 6.5, p.137) (N = 25)
Mean Range SD N number of specimens, SD standard deviation
4.98 (4.65–5.20)
2 Rodents
of origin or stratigraphic horizon of the specimens, because, for instance, EP 148/01 and EP 2033/03 both come from Loc. 6 and the same horizons. They are very different in size, but similar in cusp morphology. One can also observe variability in the disposition of the prelobe cusps on M/1 (Fig. 2.19). Aethomys sp. (Fig. 2.20) One large murid mandible with M/1-2 (EP 1648/00) has been recovered from Loc. 3 in the Upper Laetolil Beds between Tuffs 7 and 8 (Fig. 2.20). By its large size, the absence of a link between the prelobe and second lobe of the M/1, the existence of a large Cv5, the absence of cusplike Cp on M/1 and better developed on M/2, and the trace of a longitudinal crest on the second lobe, this specimen can unambiguously be attributed to Aethomys. It is the first record of this genus at Laetoli, but this taxon has already been identified at various East African sites, such as Olduvai Bed I (Jaeger 1976), Natron (Denys 1987b), East Turkana (Black and Krishtalka 1986), the Omo (Wesselman 1984), and Kanapoi (Manthi 2006), and it is possibly present at Lemudong’o (Manthi 2007). It is also recorded from Langebaanweg (Denys 1990) and other South African cave sites (Pocock 1987). This specimen differs from A. lavocati from Olduvai Bed I and from A. modernis of Langebaanweg by the absence of a longitudinal crest linking the prelobe to the first lobe of
Fig. 2.20 SEM images of Aethomys sp. EP 1648/00 from Loc. 3, Upper Laetolil Beds. Scale bar indicates 1 mm
35
M/1. Such a cusp prelobe disposition recalls A. deheinzelini from Omo Members F and G (Wesselman 1984). It also recalls the large A. adamanticola from Langebaanweg, with the same cusp prelobe disposition (no link of the prelobe and first lobe, no tmA, presence of cv1 and cv5, a small Cp on M/1 and the presence of a strong cv1 and cv5 on M/2). Among the modern Aethomys species, East African forms of the A. kaiseri-hindei group correspond in morphology to the Laetoli specimens (with less developed cingular margin and cusps), but not the modern A. chrysophilus, which displays a longitudinal link on the M/1, or A. namaquensis, which has a tma. The M/1 dimensions of Aethomys sp. from Laetoli (2.33 × 1.52 mm) are much smaller than those of A. adamanticola (ranging between 3.00 and 3.15 mm long), A. deheinzelini and A. lavocati (greater than 2.5 mm in length) (see data in Jaeger 1976; Wesselman 1984; Denys 1990). Comparisons cannot be made with the Lemudong’o and Kanapoi specimens, which only have the M1/ figured and measured (Manthi 2006, 2007). The scarcity of the Pliocene remains attributed to Aethomys prevents further identification of the Laetoli Aethomys to a known or to a new species. Mastomys cinereus Denys, 1987 (Fig. 2.21) In addition to a large Aethomys and a medium-small Thallomys laetolilensis, there is a small Murinae from Locs. 8 and 11 (from between Tuffs 7 and 8), represented by a right mandible with M/1-2 (EP 1485/03 from Loc. 8) and a mandible with M/2 (EP 2592/00 from Loc.11) (Fig. 2.21). The M/1 is very worn and broken, but displays the remains of two
Fig. 2.21 Mastomys cinereus from the Upper Laetolil Beds. (a) EP 1485/03 (Loc. 8), left mandible with M/1-2; (b) EP 2592/00 (Loc. 11), left mandible with M/2 and roots of M/1 and M/3
36
C. Denys
anterior cusps, relatively well fused cusps with poorly marked synclinals, a small posterior cingulum, and the absence of a cingular margin that all characterize the species (Plate 6.4 in Denys 1987a). This molar measures 1.76 × 1.00 mm, which corresponds to the dimensions of the M. cinereus material of the previous collections (Denys 1987a: table 6.6). The M/2 of the specimen EP 1485/03 is also very worn and has nearly the same size as that of EP 2592/00 (1.29 × 1.05 mm and 1.23 × 1.05 mm respectively). The less worn M/2 displays a cv1 and cv5, a small posterior and two lobes with a large tC or tD (labial one), as in the isolated M/2 (79/A6108) figured in Denys (1987a: plate 6.4). Moreover, the root pattern on this mandible confirms that M/1 and M/3 each have two roots. In the previous records, Mastomys cinereus was known only from Locs. 5 and 6, and was not identified from Locs. 8 and 11 (Denys 1987a). Infraorder Hystricognathi Brandt, 1855 Family Thryonomyidae Pocock, 1922 Genus Thryonomys Fitzinger, 1867 Some hystricognathous mandibles and four-lophed isolated molars were attributed by Denys (1987a) to Thryonomys sp. They all came from the Upper Ndolanya Beds (Locs. 7E and 18). The new material allows refinement of the descriptions and a better assessment of the relationship to modern and fossil representative.
Fig. 2.22 SEM images of Thryonomys wesselmani sp. nov. from the Upper Ndolanya Beds. Left, EP 1324/05 (holotype) from Loc. 22S (P4/M3/). Right, EP 1251/00 from Loc. 22S (left mandible with M/1-2)
Thryonomys wesselmani sp. nov. (Figs. 2.22–2.25) Holotype: EP 1324/05, maxillary fragment with P4-M3/ (Figs. 2.22 and 2.24) Paratypes: EP 814/01, right mandible with M/1-3 (Fig. 2.25), associated with left mandible with M/2-3, Loc. 18. EP 464/05, left mandible with M/2-3 and incisor, Loc. 18. EP 1251/00, left mandible with M/1 and broken M/2 (very damaged), Loc. 22S (Fig. 2.22). EP 1252/00, isolated upper incisors, Loc. 22S (Fig. 2.23). Type locality: Loc. 22S, Upper Ndolanya Beds, Laetoli. Age and horizon: Mid-Pliocene (2.66 Ma), Upper Ndolanya Beds at Laetoli Locs. 7E, 18 and 22S. Etymology: in honor of Hank Wesselman who described the Omo rodents. Referred material: LAET 76-32 DP/4 (Fig. 2.25), LAET 76-700, M/1-2 (Plate 6.8 in Denys 1987a), LAET 76-117, DP/4-M/2, LAET 73-73, DP4-M1/ (Plate 6.8 in Denys 1987a). Diagnosis: Intermediate-sized Thryonomys, smaller than T. swinderianus and slightly larger than T. gregorianus. It has a less straight lingual part of the posteroloph on M1/, more elongated DP/4, low crowns and roots visible. The DP/4 has four lophs and is narrow. The upper incisor exhibits three grooves, not equally distributed along the buccal surface of the incisors as in T. gregorianus, but grouped on the lingual side of the incisor as in T. swinderianus (Fig. 2.23). Distinct from the
Fig. 2.23 Buccal side of the upper incisor of Thryonomys wesselmani sp. nov. (EP 1251/00). Scale bar indicates 1 mm
extinct Miocene Paraphiomys in lacking a mesoloph (Thryonomys has only three lophs on the upper molars) and relatively similar to fossil Paraulacodus and modern Thryonomys species.
2 Rodents
Fig. 2.24 Drawings of upper molars of T. wesselmani sp. nov. compared with other fossil and modern Thryonomys spp. (a) holotype, M1/, T. wesselmani sp. nov.; (b) LAET 73A, M1/, T. wesselmani sp. nov.; (c) right M1/ Thryonomys gregorianus Omo L1-374, Member B;
37
(d) left M1/ of a young modern T. gregorianus, BM(NH) 32.864, Kenya; (e) M1-2/ of modern T. swinderianus, MNHN 1892-1608 from the Congo
upper incisors and by its larger size. It is distinguished from the Middle Awash late Miocene Thryonomys asakomae (Wesselman et al. 2009) by the presence of three grooves on the upper incisors instead of two and by its larger size. Measurements: Tables 2.10 and 2.11
Fig. 2.25 Lower molar row of T. wesselmani sp. nov. (a) EP 814/01, right mandible with M/1-3; (b) LAET 7E-32, right DP/4
Differs from the modern T. swinderianus by the lower hypsodonty (lower crowns with visible roots), and smaller size. Differs from T. gregorianus by the groove disposition on the upper incisor and larger size of the molars. It is distinct from Paraulacaudus by having less oblique lophs on the M/1, the presence of three versus two grooves on the
Description: The holotype has quite broken and worn molars, but on these one can distinguish the presence of three lophs on the upper molars as in modern Thryonomys (Figs. 2.22 and 2.24). The crowns are very low. The upper incisors are much narrower than in modern Thryonomys (Table 2.11) and display three grooves placed along the lingual half of the buccal surface (two large grooves of equal size and a small one, Fig. 2.23). This disposition is similar to T. swinderianus incisors, which also display three grooves grouped on the internal part of the surface, and the first groove is much deeper than in the fossil (figured in Kingdon 1974). The DP4/ of T. wesselmani is in a bad state of preservation and none was previously recorded in the old Laetoli collections, but one can distinguish three lophs on a small, squared crown with an anteroloph running obliquely toward the anterior wall of the tooth and three oblique parallel lophs. The posteroloph (loph III of Denys 1987a) is relatively transverse and long, not convex distally. On the upper dental row, the protocone and hypocone are large, the hypocone being much more developed than the protocone, and they are relatively transverse as in modern Thryonomys spp. The labial valley (or sinus) separating these two cusps is oblique as in modern species. On the M1/ and M2/ of the holotype there is a little
18.48
5.32 × 3.73 5.65 × 3.64 5.84 × 4.70 5.40 × 4.10
6.15 × 4.56
4.83 × 3.27
(4.8) × 3.1 4.8 × 3.6
4.9 × 3.1
4.23 × 4.29 4.45 × 4.46 4.50 × 4.94 4.60 × 4.92
1.50 × 1.20 2.00 × 1.70
4.61 × 4.53 3.25–4.24 × 3.40–4.30
4.7 × 4.8 4.2 × 4.8
5.00 × 4.66
4.55 × 3.85
4.31 × 4.62 4.45 × 4.31 4.94 × 4.72 4.94 × 5.36
1.50 × 1.40
4.85 × 4.57 4.00–4.10 × 3.73–4.27
4.4 × 5.2
4.89 × 4.17 4.81 × 4.96
4.35 × 3.79
4.62 × 4.37 4.14–4.32 × 3.84–4.07
4.71 × 4.18 4.52 × 4.56
15.55
15.45+
3.70 × 4.28
3.07 × 3.79
3.56–3.63 × 4.92–4.93
3.79 × 4.53
4.12 × 4.84 4.14 × 4.90 4.64 × 5.00 4.16 × 5.4
3.22 × 3.44
3.18–3.89 × 3.44–5.13
3.89 × 5.9
4.90 × 5.00 4.20 × 5.8
4.36 × 4.22
5.60 × 6.50
3.14–3.69 × 3.73-4.61
4.32 × 6
M3/
T. swinderianus 1892-1607 Central African Republic 20.27 4.56 × – 4.62 × 5.76 5.23 × 5.70 17.20 4.87 × 6.56 5.66 × 6.4 5.59 × 6.49 1974-35 Gabon 20.85 6.42 × 4.18 4.66 × 5.45 4.76 × 5.49 4.75 × 5.16 18.03 4.62 × 5.95 4.58 × 6.25 5.83 × 6.05 1991-200 Congo 20.62 6.30 × 4.22 4.75 × 5.47 4.77 × 5.14 4.59 × 5.54 18.19 3.97 × 5.53 4.46 × 5.44 4.52 × 5.17 Data sources: Manonga, Thryonomys sp. (Winkler 1997), T. asakomae (Wesselman et al. 2009), Lemudong’o (Manthi 2007), Omo Shungura, T. swinderianus and T. gregorianus (Wesselman 1984), modern representatives from museum collections. LTR: Upper tooth row length (DP/4-M/3); UTR: Upper tooth row length (DP4/-M3/)
T. gregorianus OM6435 Kenya OM7595 Kenya BM 30.3.4.4 Kenya BM 32.8.6.4 Kenya
T. gregorianus Omo B
T. swinderianus Omo J
Thryonomys sp. Manonga Lemudong’o 45945 Lemudong’o 45934
Mean T. asakomae
T. wesselmani EP 1251/00 EP 464/05 EP 814/01 EP 1324/05 LAET 74-32 LAET 75-700 LAET 75-117 LAET 74-31
Table 2.10 Tooth dimensions (mm) for fossil and modern Thryonomys spp. Thryonomys wesselmani includes material from this work and Denys (1987a) Specimen LTR DP/4 M/1 M/2 M/3 UTR M1/ M2/
38 C. Denys
2 Rodents
39
Table 2.11 Buccolingual width of the upper (UI) and lower (LI) incisors of modern Thryonomys swinderianus from MNHN collections and of fossil T. wesselmani sp. nov Species UI width LI width T. swinderianus 1892-1607 T. swinderianus 1947-35 T. swinderianus 1991-200 T. wesselmani sp. nov. EP1252/00 EP 464/05 EP 814/01 Omo J Omo C Omo F
5.79 5.42 5.82
5.72 5.17 5.61
4.77 5.31 5.65 3.67 4.88 4.40
inflexion of the loph at the place where the anteroloph starts on P4/ and is reminiscent of this crest (Fig. 2.22). This is not visible on the modern Thryonomys swinderianus, which have very rectilinear lophs (Fig. 2.23). The M1/ or M2/ was described in Denys (1987a: fig. 2, plate 6.8), and we summarize here the main features. It bears three lophs. The anteroloph is long and convex and joins a small crestiform protocone at the anterolingual corner of the tooth. The metaloph is oblique and is prolonged by the paracone, which is situated in the anterolingual part of the tooth. The posteroloph is also long and reaches the posterolingual corner of the molar to a crestiform metacone, which is also nearly longitudinal in its disposition. The protocone is nearly longitudinal, while the hypocone is oblique; both are joined by a longitudinal ectoloph. The crown is low and the cusps are bunodont. The DP/4 is broken in all the new specimens, but LAET 17E-32 displays four lophs as in modern Thryonomys species (Fig. 2.25). Denys (1987a) mentioned that the M/1 protoconid and hypoconid are more transverse than in modern Thryonomys (where they are very oblique and crestiform) and the lophs are convex distally, which is also visible in EP 814/01 (Fig. 2.25), while in the modern species they are much more rectilinear and transverse. There is a short anterolophid running obliquely from the protoconid, which is visible as a separate cusp/crest on the M/1-2; a feature not seen on modern Thryonomys spp. or only present as a small inflated zone incorporated into the anterolophid. Lophid III is shorter in T. wesselmani than in modern species, where it occupies the whole breadth of the molar. The M/3 is preserved in EP 814/01 and displays three lophs and a short anteroconid incorporated into the base of the protoconid and protolophid. The molar is as long as M/2, but narrower distally, with a very reduced hypoconid and a very small posterolophid (loph III). Compared with modern Thryonomys, it is smaller, less convex and crestiform, and the distal half of the M/3 is proportionally wider, with the same proportions and width as the M/2. This loph is narrower on modern Thryonomys
only when the molars are slightly worn and there is no link between the two lobes of the molar. In EP 814/01, which is intermediate in wear, the link is made between the two distal lobes of the molars, and it has low crowns (Fig. 2.25). The oldest Thryonomys comes from late Miocene deposits of the Middle Awash of Asakoma, Biki Mali Koma, and Gigiba Dora localities, all dated at 5.7 Ma (Wesselman et al. 2009). On the figured molars of T. asakomae one can see that the lophs are much longer and more transverse, and the cusps are much more crestiform than in the new Laetoli species. The crowns appear higher, and this gives the Middle Awash fossils a very modern aspect. All the molars of T. asakomae are larger than the Lothagam and Manonga specimens, and smaller than Laetoli T. wesselmani and the modern Thryonomys species (Table 2.10). In addition, Thryonomys cf. gregorianus was described from the Nachukui Formation at Lothagam (Winkler 2003), and specimens attributed to Thryonomys sp. occur in the Manonga Valley at Ibole (Winkler 1997). They are represented by two upper molars of much smaller size and moderate hypsodonty compared to Thryonomys from the Upper Ndolanya Beds. They could belong to a different species from the Laetoli material, and may represent an ancestral form. According to Wesselman et al. (2009) they are related to modern T. gregorianus. The discovery of new specimens of a bundont Thryonomys at Laetoli confirms that the divergence between the two modern lineages of Thryonomys had already occurred by the mid-Pliocene. The only other record of the genus from East Africa is from the Omo Shungura Formation, where Wesselman (1984) recorded both T. swinderianus in Member J and T. gregorianus in Members B, C and F. Examination of these specimens shows that the Omo L1-174 (from Member B, Fig. 2.24) assigned by Wesselman (1984) to T. gregorianus belongs to a young animal. It displays a small anteroloph, not reaching the labial corner of the molar, and it is smaller in size than modern T. gregorianus (Table 2.10). It could belong to a species distinct from the modern one, and close to T. wesselmani. As for the specimens referred to T. swinderianus from Omo Member J, they are clearly larger than T. wesselmani. Manthi (2007) figured and describe a thryonomyid indet. of small size from Lemudong’o that displays three transverse lophs on the M/1-2. It is difficult to assign this specimen to any previously described species, but it could be the earliest known representative of the genus at 6 Ma (Table 2.10). No Thryonomys specimens have been described from the South African Pliocene sites. T. wesselmani sp. nov. retains some primitive characters, such as the anteroloph/anterolophid trace and its intermediate size, and it could represent the ancestor of the two species living in tropical Africa today. Compared to Thryonomys sp. from Ibole (Manonga Valley) described by Winkler
40
C. Denys
(1997), T. wesselmani has a smaller upper M1-2, even smaller than those we have measured for T. gregorianus, the smallest of the modern species. However, the paucity of the Ibole and Laetoli material does not allow refinement of species attributions, but they probably constitute two valid species. Family Petromuridae Wood, 1955 Petromus sp. A. Smith, 1831 (Figs. 2.26–2.27) Three very damaged mandibles from the Lower Laetolil Beds (EP 014/98, mandible with M/1, Kakesio; EP 014/99, left mandible with M/2 (Fig. 2.26), Kakesio; and EP 2076/03, right mandible with P/4-M/3, Emboremony 1) (Fig. 2.27), can be attributed to Petromus because the molars display only two wide lophs and traces of a small posteroloph (Fig. 2.26). The distal parts of the mandibles are broken, but there are signs of less of a hystricognath disposition than in Thryonomys, which characterizes the modern Petromus. The M/3 displays two distinct anterior cusps (or a cusp consisting of two fused ones) as in modern P. typicus from South Africa (Fig. 2.27). The presence of a very small posteroloph differentiates it also from Paraulacodus and Paraphiomys, which have three and four lophs on M/1-2. The DP/4 is clearly distinct from that of Thryonomys in displaying only three lophs and in being very simplified, with no mesolophid (Fig. 2.27). In some regards, the DP/4 resembles P. shipmani from Fort Ternan (Denys and Jaeger 1992) and P. roessneri from Harasib (Mein et al. 2000b). The Lower Laetolil specimens have a strong metaconid linked to the hypoconid by a slightly oblique crest, and there is still a trace of a posteroloph that is not present on specimens of P. antiquus from the early Pliocene of South Africa (Sénégas 2004). Petromus antiquus has an ectolophid situated in the middle of the molars and arranged longitudinally, and the cusps are much more transversely fused and aligned than in the Lower Laetolil specimens. Mein and Pickford (2006) briefly described a single left mandible with damaged M/1-2 from Kapsomin, Lukeino Formation, which they identified as P. cf. antiquus. The published photograph does not allow identification of the main characters, except for the fused and transverse labial and lingual cusps. On modern P. typicus the two lophs are oblique, the distal cusps are well fused and poorly individualized, the teeth have higher crowns and the presence of cement, and they are of comparable size or slightly smaller than the Lower Laetolil specimens (Table 2.12). The Lower Laetolil Petromus sp. clearly represents a very early evolutionary stage, and it could represent an extinct genus intermediate between Phiomys spp. and modern Petromus. However, pending further material, notably of upper molars, we attribute these fossils to Petromus sp. for the moment. If this attribution is
Fig. 2.26 Petromus sp. Detail of the M/2, EP 014/99, Emboremony 1 (Lower Laetolil Beds). Scale bar in mm
Fig. 2.27 Comparisons of lower dentition of Petromus. (a) EP 2076/03, right mandible with DP/4-M/3 of Petromus sp.; (b) juvenile mandible with DP/4-M/2 of modern P. typicus from southwest Africa (Cape museum collections, ZM119111A); (c) modern adult right mandible of P. typicus from the NHM collections. Drawings to the same scale
2 Rodents
41 Table 2.12 Dimensions (mm) of Petromus spp. from Lower Laetoli Beds compared with P. antiquus from Waypoint 160 (Sénégas, unpublished) and modern P. typicus P/4-M/3 EP 014/98 EP 2076/03 P. antiquus P. typicus BM25.1.2.219 P. typicus ZM119.111A
11.46
P/4 2.91 × 2.33 1.90 × 1.74 1.73 × 1.58 2.08 × 2.00
confirmed, it would be the second record of fossil Petromus for the Pliocene of East Africa. The first record being the poorly known P. cf. antiquus from Lukeino (Mein and Pickford 2006). Other fossil Petromus are known from South African sites, and recently Sénégas (2004) described P. antiquus from the Gauteng Province, South Africa at Waypoint 160 (close to Bolt’s Farm) of early Pliocene age. A single specimen was previously recorded from Taung: P. minor Broom, 1939 and a Petromus sp. is recorded from Namibian sites in the Otavi mountains and Kaokoland (Pickford et al. 1994). According to Sénégas (2004), P. minor is similar to P. antiquus, but there are some differences in molar proportions. Petromus typicus is found today only in western South Africa, Namibia and southwest Angola (Woods and Kilpatrick 2005). Family Bathyergidae Waterhouse, 1841 Genus Heterocephalus Rüppell, 1842 Heterocephalus manthii sp. nov. (Figs. 2.28–2.31) Holotype: Half cranium, KK 82-28 (currently on loan to the National Museum of Kenya, Nairobi, but part of the permanent collections of the National Museum of Tanzania, Dar es Salaam) (Fig. 2.28) Paratypes: (currently on loan to National Museum of Kenya, Nairobi, but part of the permanent collections of the National Museum of Tanzania, Dar es Salaam): KK 82-1, maxilla with left M13/ and right M2/ (Fig. 2.29). KK 82-43, maxillary fragment. All from Kakesio. Type locality: Kakesio, Lower Laetolil Beds, Tanzania. Age and horizon: Mid-Pliocene, Lower Laetolil Beds. Diagnosis: Small hypsodont Heterocephalus species with a long, bilobed M3/, well marked anterior and posterior depressions on M1/, presence of elongated distolingual angle on M1/. Differs from modern H. glaber in the proportions of the molars, smaller size, and greater hypsodonty. It is distinguished from H. quenstedti by the bilobed, longer M3/. It has smaller molars than H. atikoi from Omo Members F and G,
M/1
M/2
M/3
2.72 × 2.72 2.82 × 2.65 2.19 × 1.77 2.40 × 2.24
3.10 × 2.71 3.00 × 2.48 2.65 × 2.88 2.12 × 1.89 2.72 × 2.12
2.52 × 2.24 2.89 × 2.53 2.23 × 1.92
and H. jaegeri from Olduvai Bed I. It is less hypsodont than H. quenstedti and H. jaegeri, but it is much more hypsodont than modern H. glaber. Measurements: Tables 2.13 and 2.14 Description: One cranium with the lower jaw in articulation (KK 82-28) and two fragments of maxilla (KK 82-43 and KK 82-1) with three upper molars can be attributed to the genus Heterocephalus. These were all collected in 1982 by Mary Leakey’s expedition working in the Lower Laetolil Beds at Kakesio. To our knowledge this is the oldest representative of the genus in Africa. The anterior part of the cranium (nasal and upper incisors) is broken and the dorsal view does not show any significant differences from modern or other fossil Heterocephalus species (Fig. 2.28). The width of the interorbital constriction measures 6.4 mm, which is wider than in modern H. glaber (5.2–6.2 mm) and similar to H. quenstedti from the Upper Laetolil Beds (the holotype measures 6.2 mm) and narrower than H. jaegeri from Olduvai Bed I (6.6 mm). Neither the tympanic bullae nor the distal part of the cranium are visible, which prevents further comparison with other species. The Kakesio molar dimensions (except M3/ length) are smaller than or equivalent in size to H. quenstedti specimens from the Upper Laetolil Beds (Tables 2.13 and 2.14). The very long, but narrow M3/ of H. manthii allows it to be distinguished from H. quenstedti, as well as from other species of Heterocephalus. The M1/ is square and it presents a small anterior median depression, a very shallow labial sinus, barely extended along the labial wall of the crown (Fig. 2.29). The distolingual angle is elongated. There is a small posterior median depression. The M2/ is only slightly larger than M1/. The labial sinus is as deep as one half of the width of the tooth (Fig. 2.29). The distolingual angle is not as well marked as on M1/. There is no anterior median depression. The M3/ is the largest tooth of the molar row, but is narrower. It displays two separate lobes (Fig. 2.29). The anterior lobe is twice as wide as the posterior one, and it presents an anterior depression that is slightly lingually displaced. The second lobe is rounded.
42
C. Denys
Fig. 2.28 Heterocephalus manthii sp. nov KK 82-1 (right) from Kakesio. EP 2205/03 (left), H. quenstedti from Loc. 7, Upper Laetolil Beds
Fig. 2.29 Comparison of the upper left molar rows of Heterocephalus spp. (a) H. manthii, holotype, KK 82-1; (b) H. quenstedti, LAET 75-2808; (c) H. jaegeri, Olduvai Bed I, FLK N1 M3; (d) H. glaber (MNHN 1901-72)
This molar is as long as one of the specimens from Olduvai and the Upper Laetolil Beds, but much narrower (Table 2.14). Heterocephalus manthii displays greater hypsodonty than modern H. glaber, but a lower degree of hypsodonty than the Upper Laetolil H. quenstedti. Compared with the material from Upper Laetolil and Olduvai, the Kakesio specimens exhibit several distinctive characteristics, including the longer, bilobed M3/, and the
well-marked anterior and posterior depressions on M1/. Like H. jaegeri and H. quenstedti, one sees the elongated distolingual angle, and the absence of labial sinus on M1/ related with increased hypsodonty. This indicates that they belong to the same extinct lineage of naked mole rats. Heterocephalus manthii from Kakesio has small molars compared to H. quenstedti, but the range of variability of this species is not yet known (Table 2.14).
2 Rodents
43
Fig. 2.30 Lower molar rows of fossil and modern Heterocephalus spp. (a) left mandible with M/1-3, EP 1059/98 from Loc. 9S, H. quenstedti; (b) right mandible with M/1-3, EP 2921/00 H. quenstedti from Loc.
10E H. quenstedti; (c) right mandible with M/1-3 of H. jaegeri, FLKN1 M4 Olduvai Bed I; (d) right mandible with M/1-3 of modern H. glaber (MNHN-1978-268) from Somalia. Drawings to the same scale
Heterocephalus quenstedti Dietrich, 1942 (Figs. 2.28–2.32)
than that of the Olduvai Bed I H. jaegeri and modern H. glaber. Heterocephalus atikoi (Omo Shungura F, G) is intermediate between the largest individuals in the H. quenstedti range and the smallest ones of H. jaegeri. The M1/ of H. manthii (Lower Laetolil Beds) falls in the middle of the H. quenstedti range, while the Olduvai H. jaegeri and H. glaber have longer, but not wider molars.
Naked mole rats were relatively common in the Upper Laetolil Beds and are well represented by cranial fragments (Denys 1987a). The new collections have led to the recovery of additional specimens of this species from various horizons and localities. The specimens exhibit the same dental pattern as previously described. The new material is listed in Appendix 2.4. The following three specimens have been illustrated using SEM: EP 4151/00, left mandible with M/1-3 (Fig. 2.32), Loc. 8; EP 1082/03, right mandible with M/1-3, (Fig. 2.32) Loc.10W; EP 1990/00, half cranium with M1-2/ (Fig. 2.32), Loc. 5. Additional specimens (32 new specimens) have been recovered from Locs. 1NW, 2, 5, 7, 8, 9S, 10, 10W, 10E, 22 and 22E from levels throughout the Upper Laetolil Beds. This extends the distribution of the species at Laetoli and also increases the number of available specimens. Previously it was described from Garusi by Dietrich (1942) and from Locs. 2, 5, 6, 8, 10, 11, 13 and 21 by Denys (1987a). The length of the complete lower molar row of the three new specimens has an average of 3.95 mm (Table 2.13), which fits with the Garusi, Deturi 160, LAET 75-608 and LAET 75-3512 specimens of H. quenstedti. One new specimen displays an M1-3/ length of 4.05, which is larger than in the type specimen (LAET 75-2808, M1-3/ = 3.6 mm), but identical to LAET 76-4166, which has a length of about 4.0 mm (Table 2.13). A scatterplot of molar size for the two most common molars of Heterocephalus spp. is presented in Fig. 2.31. The M/2 of H. quenstedti from the Upper Laetolil Beds is smaller
Family Hystricidae G. Fischer, 1817 This family is quite well represented at Laetoli, with three different species of porcupines. The new collections confirm the remarkable diversity of this group in East Africa during the Pliocene. Hystrix leakeyi Denys, 1987 (Fig. 2.33) New specimens: EP 392/98, germ of P/4, Loc. 10E. EP 1037/05, germ of M/1-2, Loc. 2. EP 655/05, mandible with P/4-M/2, Loc. 6 (Fig. 2.33). EP 1377/00, left mandible with M/1, Loc. 6. EP 3068/00, maxilla with DP4-M3/, Loc. 5. EP 142/05, isolated left DP/4, Loc. 8. Referred material: Hadar and Laetoli type series in Denys (1987a). Measurements: Table 2.15. Denys (1987a) indicated that H. leakeyi occurred in Locs. 3, 5, 7, 9, 11, 12S, 15 and 20 of the Upper Laetolil Beds. The new material from Laetoli adds Locs. 6, 8, and 10E to the distribution of this species. Some new specimens display the same size and shape (i.e., small, wide, very rounded and low-crowned molars) as in Hystrix leakeyi from the Upper
44
C. Denys Table 2.13 Upper and lower molar row lengths of modern and fossil Heterocephalus spp.
1,9 1,8 1,7
Width
1,6 1,5 1,4 1,3 1,2 1,1 1,2
1,1
1,3
1,4
1,5
1,6
Length Om0 F
Om0 G
Laetolil Beds
OLD Bed I M4
OLD Bed I M5
H. glaber
OLD Bed I M3
1,6
Specimen
Species
KK 1 EP 043/01 EP 1082/03 EP 2921/00 EP 4151/00 LAET 75-608 LAET 76-4166 LAET 75-2808 DET 160 M3BED1 M4BED1 M4BED1 M5BED1 M5BED1 M5BED1 1884-1572 BM51-703
H. manthii H. quenstedti H. quenstedti H. quenstedti H. quenstedti H. quenstedti H. quenstedti H. quenstedti H. quenstedti H. jaegeri H. jaegeri H. jaegeri H. jaegeri H. jaegeri H. jaegeri H. glaber H. glaber
M/13
M13/ 3.41 3.79 4.05
3.94 3.86 3.57 3.60 3.97 3.92 4.05 3.65 3.82 3.95 3.78 3.65 3.84
4.00 3.60
3.85 3.56
1,5 1,4
Width
1,3 1,2 1,1 1 0,9 0,8
0,9
1
1,1
1,2
1,3
1,4
1,5
Length Olduvai Bedl
Kakesio Beds
Laetoli Beds
H. glaber
Fig. 2.31 Scatterplots of length × width (mm) of M/2 (top) and M1/ (bottom) of fossil and modern Heterocephalus spp. Measurements for Kakesio and Upper Laetolil Beds (this work, Denys 1987a), Omo F and G (Wesselman 1984), Olduvai Bed FLKN M3, M4, M5, after (Denys 1989b), modern H. glaber from Kenya and Ethiopia measured from NHM and MNHN specimens
Laetolil Beds and Hadar. Hlusko (2007) described a small Hystrix sp., based upon a very worn isolated molar from Lemudong’o, which displays the brachyodont pattern of H. leakeyi. The length of M/1 or M/2 is smaller (6.7 mm) than in H. leakeyi specimens from Laetoli. Hystrix sp. has been recorded at Lothagam and Lukeino. Both are of small size (Table 2.15), with visible roots and brachyodont molars that could be attributed to H. leakeyi (Winkler 2003; Mein and Pickford 2006). A third Kenyan Hystrix sp. of unknown age is recorded by Manthi (2006), and is also characterized by small molar size (Table 2.15). However, the variability of these late Miocene/early Pliocene fossils is poorly known. Hystrix makapanensis Greenwood, 1958 This larger species of Hystrix appears to be very common in East and South Africa during Plio-Pleistocene time, and it
has recently been suggested that Hystrix sp. from the Middle Awash is related to this Pliocene species (Wesselman et al. 2009). It has high crowned and rather large molars (larger than modern H. cristata and H. africaeaustralis). Referred material: EP 1996/00, isolated right M/1 or M/2, Loc. 5, Upper Laetolil Beds. EP 329/00, six associated molars, Loc. 8, Upper Laetolil Beds. EP 988/00, left mandible with DP/4-M/2, Loc. 18, Upper Ndolanya Beds. EP 086/03, isolated M1/or M2/, Loc. 18, Upper Ndolanya Beds. EP 3354/00, left lower DP/4, Loc. 15, Upper Laetolil Beds. EP2015/00 isolated right M/1-2, Loc. 5, Upper Laetolil Beds. EP 376/05, broken molar, Loc. 15, Upper Ndolanya Beds. Measurements: Table 2.15. Denys (1987a) indicated that H. makapanensis occurred in Locs. 3 and 10 of the Upper Laetolil Beds. The new material from Laetoli adds Locs. 5, 8 and 15 to the distribution of this species. The Upper Ndolanya Beds did not yield any hystricids from the Mary Leakey collections, but Dietrich recorded several specimens that were attributed to Hystrix sp. by Denys (1987a). These are from Garusi and were attributed to the Upper Ndolanya Beds. New specimens collected since 1998 come from the Upper Ndolanya beds (Locs. 15 and 18). These new findings are the first testimony of Hystrix makapanensis discovered in situ in the Upper Ndolanya Beds since 1939. Xenohystrix crassidens Greenwood, 1955 An isolated DP/4 with a very wide crown (11.88 × 8.69 mm) (Table 2.15) and displaying the same root pattern as specimens of the extinct species X. crassidens. This tooth (EP 1786/00) comes from Loc. 2 in the Upper Laetolil beds. Two other large indeterminate molars, which are very corroded,
Upper Laetolil Beds Denys (1987a)
Upper Laetolil Beds this work
Omo F Omo G Olduvai Bed I
Upper Laetolil Beds Denys (1987a)
Upper Laetolil Beds this work
Omo F Omo G Olduvai Bed I
H. quenstedti
H. quenstedti
H. atikoi
H. jaegeri
H. quenstedti
H. quenstedti
H. atikoi
M/1 M/2 M/3 M/1 M/2 M/3 M/2 M/2 M/1 M/2 M/3
Molars M1/ M2/ M3/ M1/ M2/ M3/ M1/ M2/ M3/ M3/ M2/ M1/ M2/
N number of molars, SD standard deviation Data sources: H. atikoi Omo Shungura (Wesselman 1984), H. jaegeri (Denys 1989b)
H. jaegeri
Level Kakesio (Lower Laetolil Beds)
Species H. manthii sp. nov.
Table 2.14 Molar dimensions of fossil and modern Heterocephalus spp.
2 3 1 4 1 1 1 2 8 6 1 5 4 3 4 6 5 1 2 30 45 17
Length N 1
1.21 1.36 1.52
1.21 1.32 1.18 1.04 1.22 1.43 1.02 1.17 1.31 1.24
Mean 1.08 1.14 1.19 1.18 1.4 1.03 1.12 1.24 0.95 1.1
0.06 0.09 0.07
0.13 0.09 0.18 0.04 0.02 0.04
0.05 0.11
0.06
0.13 0.17
SD
1.18–1.22 1.15–1.44 1.18–1.62 1.4–1.62
0.87–1.13 1.1–1.35 1.4–1.6 0.95–1.13 1.1–1.23 1.19–1.43
1.34–1.47 1.15–1.3 1.17–1.4
1.05–1.19
1.07–1.26 1.23–1.56
Range
3 3 1 4 1 1 1 2 8 6 1 5 4 3 4 6 5 1 2 30 45 17
Width N 1
1.18 1.66 1.61
1.47 1.63 1.2 1.03 1.4 1.28 1.08 1.31 1.29 1.44
Mean 1.02 1.17 0.91 1.28 1.34 1.13 1.12 1.19 1.05 1.47
0.08 0.08 0.07
0.08 0.08 0.1 0.06 0.05 0.06
0.06 0.08
0.02
0.04 0.03
SD
1.42–1.51 1.0–1.45 1.5–1.85 1.5–1.75
0.95–1.07 1.28–1.6 1.25–1.4 0.95–1.25 1.14–1.48 1.12–1.48
1.4–1.55 1.52–1.75
1.39–1.52
1.1–1.14
1.15–1.3 1.33–1.38
Range
2 Rodents 45
46
C. Denys
also belong to Xenohystrix (EP 3624/00 and EP 3623/00, Loc. 21, Upper Laetolil Beds). Denys (1987a) recorded Xenohystrix in the Upper Laetolil Beds at Locs. 1, 2, 10 and 15, and we add here Loc. 21 to that list. Hlusko (2007) described Xenohystrix sp. from Lemudong’o. The low crown and occlusal pattern fits well with X. crassidens from the Upper Laetolil Beds and suggests conspecificity, as suggested by Hlusko (2007). Wesselman et al. (2009) also described Xenohystrix sp. from the late Miocene Adu Dora sites and mentioned its presence at Aramis at around 3.4 Ma (Wesselman and Black, unpublished). Xenohystrix crassidens is also known from Makapansgat and Hadar, but it has never been recorded from Pleistocene sites. Hystrix sp. 1
Fig. 2.32 SEM images of Heterocephalus quenstedti molars. Left, EP 4151/00, left mandible with M/1-3 from Loc. 8; right, EP 1082/03, right mandible with M/1-3 from Loc. 10W; bottom, EP 1990/00, left maxillary fragment with M1-2/ from Loc. 5. Scale bar in mm
A single specimen (EP 2352/03), comprising two associated mandibles from the late Pleistocene Upper Ngaloba Beds (Loc. 13), exhibits a wide tooth row, with a length greater than modern H. cristata from Kenya (even old adults) (Table 2.15). The indication in the literature that H. africaeaustralis has larger teeth than H. cristata can be found in Denys (1987a: fig. 6.20). However, museum specimens are in general not well identified taxonomically, and there is great variability depending on age, with quite late replacement of DP/4 by P/4 and emergence of M3/3. Although our reference sample is too small to reach a taxonomic conclusion, this specimen does seem close in morphology to modern Hystrix species.
Discussion and Conclusion
Fig. 2.33 Hystrix leakeyi. EP 655/05, left mandible with DP/4-M/2. Scale bar indicates 5 mm
The new collections from Laetoli have allowed a more refined assessment of the status of the rodent taxa previously unattributed to species, as well as the description of several new species, including Gerbilliscus satimani, G. winkleri, Thryonomys wesselmani, Paraxerus meini and Heterocephalus manthii. It has also allowed a better appreciation of the intraspecific variability among S. major, X. janenschi, H. quenstedti and P. laetoliensis populations. Moreover, thanks to the newly recovered material, we have been able to assign Laetoli specimens for the first time to Petromus sp and Aethomys sp. Muridae are quite rare at Laetoli and only a few new remains can be attributed to this family. The new material provides additional T. laetolilensis molars and a few attributable to M. cinereus from the Upper Laetolil Beds. Unfortunately, we do not have any supplementary specimens of T. cf. laetolilensis and M. cf. cinereus from the Upper Ndolanya Beds. Since very few species are found in both the
2 Rodents
47
Table 2.15 Molar row length and lower tooth dimensions (length × width, mm) of Hystrix spp., including data from Denys (1987a), Greenwood (1955), Mein and Pickford (2006), Winkler (2003) and Manthi (2006). Measurements on modern H. cristata specimens from collections of the Kenya National Museum Number
LM/1-3
EP 2352/03
37.33
DP/4
M/1
9.14 × 6.65
EP 392/98 EP 655/05 EP 142/05 EP 1037/05 LAET 75-1368 LAET 75-2594 LAET 74-398
6.87 × 5.13 10.41 × 8.24 8.09 × 7.02 8.64 × 7.34 9.70 × 7.80
Makapan 1996 2015 3354 329 988 Olduvai LAET 75-1971 1786 LAET 75-3411 OM5329 OM5324 OM5322 OM7190 OM7114
M/2
M/3
11.44 × 9.16 6.12 × 5.90 6.6 × 5.00
6.00 × 5.90
7.05 × 6.20
9.05 × 7.87
10.13 × 9.18
7.31 × 6.06 7.28 × 7.64
7.96 × 6.70
7.34 × 6.00
8.20 × 7.10
8.7 × 7.00
7.2 × 6.10
13 × 9.05 9.64 × 8.69 10.35 × 7.10 8.99 × 7.38 10.86 × 9.51 11.5 × 9.6 12.6 × 9.5
34.95 33.44 33.62 34.40
11.88 × 8.69 14.6 × 10.4 8.69 × 5.87 8.14 × 6.91 8.65 × 5.26 10.2 × 7.54
12.13 × 9.11 11.5 × 8.9
7.43 × 6.5 8.39 × 7.26 8.35 × 5.62 7.93 × 7.02
Upper Laetolil Beds and Upper Ndolanya Beds it would have been interesting to examine the relationships between samples from these two stratigraphic units. We did not find any additional specimen of the unattributed murid indet. from the Laetolil Beds (Denys 1987a), which may be close to Acomys. The absence of Deomyinae at Laetoli is surprising, because representatives of this taxon are quite abundant in the late Miocene and early Pliocene of Ethiopia, Kenya (Winkler 1997; Geraads 2001; Manthi 2006; Mein and Pickford 2006; Wesselman, personal communication) and Harasib, Namibia (Mein et al. 2004). A recent molecular phylogeny has demonstrated the monophyly of the African Murinae and the existence of four major clades in tropical Africa (Lecompte et al. 2008), which differentiated as early as 7–8 Ma. The first Arvicanthini clade includes Thallomys, as well as Aethomys. The second one is composed of the tribe Praomyni, which comprises Mastomys. The two remaining major murine lineages, containing the
11.6 × 9.5
8.68 × 6.06 8.92 × 7.25 9.05 × 5.58 8.87 × 7.40
10.8 × 8.4
7.32 × 5.45 7.62 × 4.55 8.28 × 7.15
Species H. sp. Ngaloba Beds Hystrix sp. Lothagam Hystrix sp. Lukeino Hystrix sp. Kanapoi H. leakeyi H. leakeyi H. leakeyi H. leakeyi H. leakeyi H. leakeyi Type H. leakeyi H. makapanensis H. makapanensis H. makapanensis H. makapanensis H. makapanensis H. makapanensis H. makapanensis H. makapanensis X. crassidens X. crassidens H. cristata H. cristata H. cristata H. cristata H. cristata
Mus (Nannomys) and Malacomys clades respectively, are not yet recorded at Laetoli, despite relatively good fossil samples. The Laetolil Beds still represent the first occurrence of Thallomys. In the present study, a new murid belonging to the genus Aethomys is described for the first time from Laetoli. This taxon has already been identified at Kanapoi (Manthi 2006), Lemudong’o (Manthi 2007), Langebaanweg, Makapansgat and Bolt’s Farm (Denys 1999). Some Dendromurinae, such as Dendromus sp. and Steatomys sp. described by Denys (1987a), were not recovered again, but this may be due to sampling methods (only surface collection rather than screening) considering their very small size. Such a bias in sampling may also explain the absence of Mus (Nannomys). Manthi (2006) has recorded the presence of Steatomys sp. and Mus sp. in Nzube’s mandible site at Kanapoi (~4.1 Ma), which confirms the early Pliocene occurrence of these taxa in East Africa.
48
The new material from Laetoli allows confirmation of the absence of Heterocephalus and Aethomys in the Upper Ndolanya Beds. We do, however, confirm the presence of Hystrix makapanensis in the Upper Ndolanya Beds. The Laetolil Beds only lack Thryonomys compared with the Upper Ndolanya Beds. One should note that two Gerbillinae and Hystrix coexist in the Upper Laetolil Beds. Because Heterocephalus is a mole-rat living exclusively underground one can assume that the absence of this genus in the Upper Ndolanya Beds may result from the lack of suitable soils for tunneling. It could also result from taphonomic causes, including absence of a predator that specializes in such rodents. The absence of Thryonomys, the cane rat, in the Upper Laetolil Beds may indicate the lack of availability of leaves, stems and shoots of grasses, reeds and sedges that are an important part of the diet of these rodents, which mostly live in the high grass zones close to rivers. The Upper Ndolanya Beds would have, at least in some places, a different landscape relative to the Laetolil Beds, as previously indicated by Denys (1987a). The rodent diversity of the Upper Ndolanya Beds still remains very low. There are only nine species represented, compared with 17 species from the Upper Laetolil Beds. However, only three species are recorded from the Lower Laetolil Beds (Table 2.16). The present work has added three species and one genus to the general faunal list of the Upper Laetolil Beds and two species to the Upper Ndolanya Beds. By plotting the diversity against the number of identifiable specimens (NISP) it can be shown that there is a direct link between the high NISP and the highest diversity in the Laetolil Beds (Table 2.17). However, the diversity is much lower in the Upper Ndolanya Beds compared to the Upper Laetolil Beds. Some taphonomic considerations may allow a better understanding of these differences between the Upper Laetolil and Upper Ndolanya accumulations. This will allow an appreciation of whether there are paleoecological differences between the two rodent communities, as highlighted previously by Denys (1987a) and Gentry (1987), or whether the relatively low diversity in the Upper Ndolanya Beds results from a different mode of accumulations of fossil remains compared to the Laetolil Beds. We also observe the increased temporal range of some species, such as Pedetes, which is now known to occur in the Upper Ndolanya Beds, and X. janenschi, which now occurs in Upper Laetolil Beds. The only rodent present in the three main stratigraphic units, and displaying the same size and morphology, is Saccostomus major. This species occurs first at Ibole in the Manonga Valley at around 5–4 Ma (Winkler 1997). Since the same species is found in both the Manonga and the Lower Laetolil Beds it confirms that the two sites are close in age and belong to the same biogeographical province. Saccostomus major occurs in sites between 5–2.7 Ma and 1.7 Ma in Tanzania. Its
C. Denys Table 2.16 List of rodent taxa from the Laetolil and Upper Ndolanya Beds (based upon this work and Denys 1987a) Species Pedetes laetoliensis Pedetes sp. Xerus janenschi Xerus sp. Paraxerus meini Gerbilliscus satimani Gerbilliscus winkleri Gerbilliscus cf. inclusus Dendromus sp. Steatomys sp. Saccostomus major Aethomys sp. Murid indet. Thallomys laetolilensis Mastomys cinereus Heterocephalus quenstedti Heterocephalus manthii Thryonomys wesselmani Petromus sp. Hystrix leakeyi Hystrix makapanensis Xenohystrix crassidens Species richness
Lower Laetolil Beds
Upper Laetolil Beds
Upper Ndolanya Beds
x x x
x x x x
x
x x
x
x x x
x
x x x
cf.
x
cf.
x x x x x x
x
x 3
17
9
Table 2.17 NISP (number of identifiable specimens) of rodent genera and species recorded from the Upper Laetolil Beds and Upper Ndolanya Beds in 1987, in this work, and combined 1987 ULB UNB This work ULB UNB Total ULB UNB
NISP
Genera
Species
243 55
13 8
15 8
227 45
10 9
17 9
470 100
15 9
17 9
local disappearance, not yet documented in any intervening site after the Upper Ndolanya Beds (2.66 Ma) and before the FLKNN1 Olduvai site (1.7 Ma), may be due to the strong
2 Rodents
49
climatic event that occurred around 2.4 Ma (see references in Maslin and Christensen 2007). Concerning the Upper Laetolil Beds, we do not see any real differences between the different localities or different stratigraphic levels in relation to Saccostomus. Along with Pedetes, Saccostomus is the dominant rodent in the assemblage (75 and 78 individuals respectively based on the new material), followed by Heterocephalus (Minimum number of individuals or MNI = 32) and Thallomys (MNI = 16). In the Upper Ndolanya Beds, Xerus (MNI = 30) dominates the assemblage, followed by G. winkleri (MNI = 11) and T. wesselmani (MNI = 5). The presence of Petromus at Kakesio (Lower Laetolil Beds) could indicate the close proximity of a rocky area or relatively dry conditions. Today, the dassie rat is a southwest African endemic, but during the Pliocene they were present at Taung, Bolt’s Farm and Lukeino (Sénégas 2004; Mein and Pickford 2006). This suggests the existence of a common southern savanna biogeographic province (or southwest arid region) extending from southwest-central Africa to KenyaTanzania. This is corroborated by the common presence of Saccostomus at Harasib, Lukeino, Ibole and Laetoli, which is unique to these sites, and is not found in early Pliocene sites in the Transvaal region. Alternative hypotheses to the presence of Petromus in Laetoli are that there was an extension of the Namib desert to the northeast of Africa, allowing Petromuridae to colonize Kenya and Tanzania, or that these
rodents were abundant throughout East Africa in the Pliocene, followed by a marked reduction in their geographic distribution for unknown reasons. Petromus is rare in East and South African sites, but this may also be due to taphonomic causes. In the absence of geographically intermediate sites of Miocene age we cannot give preference to any particular hypothesis. Figure 2.34 presents an F1 × F2 graph of the correspondence analysis based on the presence-absence of rodent genera from Mio-Pleistocene sites. Axis 1 shows a clear separation of the late Miocene localities (i.e., Chorora, Harasib 3, Lukeino, Ibole), which have a high diversity of extinct thryonomyids and a low count of modern genera (Fig. 2.34). Along axis 2 there is a continuous distribution of the Lower Pliocene to Pleistocene sites, which are dominated by modern genera. On the positive part of axis 2 one sees all the southern African localities grouped together (i.e., Langebaanweg, Makapansgat, Sterkfontein, Humpata, Taung, Kromdraai, Ngamiland) and on the negative part of axis 2 one finds a grouping of East African sites, including Lemudong’o and Kanapoi. At the extremes of axis 2, one finds Langebaanweg and the Ethiopian localities of Adu Asa and Hadar. The two Laetoli faunas (Upper Laetolil and Upper Ndolanya Beds) are positioned close together and are associated with Lemudong’o, Kanapoi, Omo Shungura Members B, C, F and G, Olduvai Beds I and II, and Natron. There is
Fig. 2.34 Correspondence analysis diagram on presence-absence of the rodent genera (except hystricids) for late Miocene to late Pleistocene East African and South African sites (after Winkler et al. 2010; data taken from the literature). Site abbreviations: LBW Langebaanweg; NOS Nosib; SK Swartkrans; KB, KA Kromdraai A, B; H2 Humpata 2; JAG Jägerquelle;
STS Sterkfontein; SE Sterkfontein extension; NGA Ngamiland; NAT Natron (Peninj); BBI Olduvai Base Bed I; SBI Olduvai Upper Bed I; SII Olduvai Upper Bed II; BII Olduvai Base Bed II; OMF, OMG, OMM, OMC, OMB Omo members F, G, M, C, B; LEMUD Lemudong’o; UNB Upper Ndolanya Beds; LB Laetolil Beds; ET East Turkana
50
C. Denys
some distance between the Upper Laetolil Beds and Upper Ndolanya Beds indicating the two faunas are not very similar. These results confirm the hypothesis of a regional differentiation of the rodent faunas during the Plio-Pleistocene times, and a strong link between rodent taxa and vegetation, at least as early as 6 Ma (Denys 1985, 1999). Due to its peculiar faunal composition and its close affinities with Ibole, Olduvai Bed I, Omo, Lemudong’o and Kanapoi, the Laetoli faunas represent a distinct type of rodent community in comparison to the Ethiopian Hadar and Adu Asa sites. With a distance of 3,000 km, the Rift Valley and the Equator separating the two sites, Laetoli and Hadar may have had quite different vegetation and climates. Rodents are very restricted in terms of their habitats, while hominins and other large mammals often have a better capacity for dispersal. This study highlights the importance of rodents as a tool for paleoenvironmental and paleogeographical reconstructions. Acknowledgements Thanks to T. Harrison for providing the Laetoli rodents for study and for his attentive reading of the manuscript. I am very grateful to Hank Wesselman and Alisa Winkler for helpful comments that improved the manuscript. Thanks to Fredrick K. Manthi for the courtesy of photographing the Heterocephalus holotype from Kakesio. SEM pictures were taken by Mrs. C. Chancogne-Weber of the MNHN Palaeontology laboratory, while R. Vacant prepared some cranial specimens. Thanks to all the curators of the various Natural History Museums who allowed study of specimens in their care, especially those at the NHM London, Dr. P. Jenkins, the Durban Museum, Dr. P.J. Taylor, the South African Museum, Cape Town, Dr. M.D. Avery, Seith Eiseb in Windhoek (SMM National Museum of Namibia), staff of the Kenya National Museum in both Palaeontology and Zoology Departments (N. Munida and M.G. Leakey).
Appendix 2.1 List of new material attributed to Pedetes laetoliensis from the Upper Laetolil Beds (ULB) EP number Locality Level Anatomical element 1161/01 1089/05 1090/05 1091/05 1783/00 1867/00 1868/00 3067/00 714/00 640/03 1036/05 994/05 220/00 1993/00
1 1 1 1 2 2 2 2 2 2 2 2 1 5
ULB ULB ULB ULB ULB ULB ULB ULB ULB ULB ULB ULB ULB ULB
398/03 399/03 1334/04 1335/04
5 5 5 5
ULB ULB ULB ULB
Right mandible + P/4-M/2 Maxilla with P/4-M/3 Anterior cranium + molars Two isolated molars Astragalus Right mandible + P/4-M/3 Left mandible + M/1-2 Astragalus Left mandible + P/4-M/3 Postcrania Two isolated molars Isolated molar Isolated molar Left + right mandibles + P/4-M/2 Left mandible + M/2-3 Six isolated molars Right mandible + P/4-M/2 Four isolated molars
784/05 785/05 1392/00 1444/04 1445/04 3905/00
5 5 6 6 6 7
ULB ULB ULB ULB ULB ULB
616/05 327/00 1572/01 1420/03 1421/03 1422/03
7 8 8 8 8 8
ULB ULB ULB ULB ULB ULB
1423/03
8
ULB
1088/98 1509/98 239/99 2435/03
9 9 9 9S
ULB ULB ULB ULB
783/03 784/03
9 9
ULB ULB
992/04 1263/05 240/05 1562/98
9 9 9 10E
ULB ULB ULB ULB
234/98 255/98 257/98 383/98 745/98
10E 10E 10E 10E 10W
ULB ULB ULB ULB ULB
078/99 079/99 2914/00
10E 10E 10E
ULB ULB ULB
2916/00 2920/00
10E 10E
ULB ULB
794/00 572/01 573/01 885/03
10E 10E 10E 10E
ULB ULB ULB ULB
887/03 1264/04 046/04 048/04
10E 10E 10E 10E
ULB ULB ULB ULB
878/05 881/05 068/05 536/05 2367/03
10E 10E 11 12E 13
ULB ULB ULB ULB ULB
Six associated molars Isolated molar Left mandible+ M/1-2 Right mandible + P/4-M/2 Eight isolated molars Left mandible fragment + P/4 Two isolated molars Right mandible + P/4-M/1 Right mandible + M/2-3 Left maxilla + P4/-M1/ Six isolated molars Premaxilla + 2 incisors + postcrania Distal femur + tibia + proximal femur + pelvis Three isolated molars Right mandible + P/4-M/3 Left mandible + M/1-3 Mandible + M/1-3 aggregated together by tuff Twelve isolated molars Associated postcranial fragments Seven isolated molars Isolated molar Five isolated molars Two left mandibles + M/1-2 + M/1 Left maxilla + P4-M2/ Five isolated molars Right mandible + M/1-3 Right mandible + M/2-3 Two associated molars M/1-2 Right mandible + P/4-M/2 Eight isolated molars Cranium + incisor + P4-M3/ left + right mandibles + right P/4-M/3 Right maxilla + M2/ Vertebrae + metapodials + bone fragments Three isolated molars Six isolated molars Premaxilla + incisors Left + right mandibles + P/4-M/2 + left and right M/1-2 Right mandible + P/4-M/2 Four isolated molars Five isolated molars Cranium + molars + mandible + postcranial Ten isolated molars Partial skeleton Right mandible + P/4-M/2 Two isolated molars Partial skeleton (continued)
2 Rodents
51
Appendix 2.1 (continued) EP number Locality
Level
Anatomical element
1441/98 603/01 604/01 289/03
15 16 16 16
ULB ULB ULB ULB
290/03 190/05 3625/00 529/00 1235/98 3741/00
16 16 21 21 22 22
ULB ULB ULB ULB ULB ULB
573/00 144/04 1220/05
22 22 22E
ULB ULB ULB
Four isolated molars Right mandible + P/4-M/2 Left mandible + M/1-3 Complete femur, premaxilla, germ P4/, other fragments Left mandible + P/4-M/2 Isolated molar Right maxilla + P4/ Right mandible + P/4 Left toothrow P/4-M/3 Anterior cranial fragment + right maxilla + P4-M1/+left maxilla + P4-M2/ Left maxilla + M1-3/ Three isolated molars Isolated molar
Appendix 2.3 List of specimens attributed to Saccostomus major from the Upper Laetolil Beds (ULB) and S.cf. major from the Upper Ndolanya Beds (UNB) EP number Locality Level Anatomical element 218/00 1162/01 1890/03 1878/00 1879/00 4255/00 4256/00 732/00 584/03 651/03 652/03 1738/04 1038/05 997/05 998/05 2745/00 2746/00 226/01 508/03 659/04
1 1 1 2 2 2 2 2 2 2 2 2 2 2 2 3 3 3 3 3
ULB ULB ULB ULB ULB ULB ULB ULB ULB ULB ULB ULB ULB ULB ULB ULB ULB ULB ULB ULB
Appendix 2.2 List of specimens attributed to Xerus janenschi from the Upper Laetolil Beds (ULB) and the Upper Ndolanya Beds (UNB) EP number Locality Level Anatomical element
185/03 186/03 1987/00
4 4 5
ULB ULB ULB
1534/01 1561/01 2512/03 033/03 034/03 1134/05 1135/05 1136/05 3648/00
Silal Artum Silal Artum Silal Artum Silal Artum Silal Artum Silal Artum Silal Artum Silal Artum 9S
UNB UNB UNB UNB UNB UNB UNB UNB ULB
3496/00
15
UNB
3497/00 3498/00
15 15
UNB UNB
1988/00 1989/00 3071/00 3072/00 392/03 393/03 394/03 395/03 395/03 1331/04 782/05 783/05 1375/00
5 5 5 5 5 5 5 5 5 5 5 5 6
ULB ULB ULB ULB ULB ULB ULB ULB ULB ULB ULB ULB ULB
3499/00 4055/00 1690/03 1691/03 1692/03 1699/03 219/04 383/05 1000/00 2356/00 813/01 816/01
15 15 15 15 15 15 15 15 18 18 18 18
UNB UNB UNB UNB UNB UNB UNB UNB UNB UNB UNB UNB
1455/04 3904/00 1247/03 1248/03
6 7 7E 7E
ULB ULB UNB UNB
087/03 292/04 293/04
18 18 18
UNB UNB UNB
1249/03 1958/03 2138/03 2211/03 1424/03 1425/03 1426/03 361/04
7E 7 7 7 8 8 8 8
UNB ULB ULB ULB ULB ULB ULB ULB
Neurocranium Mandible + P/4-M/1 Cranium Skull Left mandible + M/1-M/3 Anterior cranium Left mandible + P/4-M/3 Left mandible + P/4-M/3 Left mandible + M/1-M/3+ partial skeleton Maxilla fragment + M/1 + postcranial Partial skull Anterior cranial fragment (edentulous) Left mandible M/1-M/2 Right mandible + P/4-M/2 Cranium Left mandible + P/4-M/3 Right mandible Molar Cranium Cranium Right mandible + P/4-M/3 Maxilla + P4-M3 Right mandible + P/4-M/3 Left mandible + M/1 broken + M/2 Right maxilla + M1-2/ Mandible + P/4-M/3 Mandible + P/4-M/3
Left mandible + M/2-3 Right mandible + M/1-2 Left maxilla + M1-2/ Left mandible + M/1-3 Right mandible + M /1-3 Right mandible with M/1-2 Left mandible + M/2 Left mandible + M/1-2 Left maxilla + M1/ Right mandible + M/1 Right maxilla + M1-2/ Left mandible + M/1-3 Left mandible + M/1-2 Right maxilla + M1/ Left maxilla + M1-2/ Right mandible + M/1-2 Right mandible + M/2-3 Right mandible + M/1 Right maxilla + M1-2/ Cranial fragment + right M/1-3 Left mandible + M/1-2 Left maxilla + M1-2/ Left mandible + M/3 and right mandible + M/1-3 Right mandible + M/1-3 Right mandible + M/2-3 2M1/+2M/1 + postcranials 2 incisors Right M1/ Right maxilla + M1-3/ Left mandible + M/2 Right mandible + M/2 Right mandible + M/2-3 Left mandible + M/1-2 Right mandible + M/1-3 Left maxilla + M1-2/ Associated left maxilla + M1/ and right maxilla + M1-2/ and right mandible + M/1-3 Right mandible + M/1-2 Right maxilla + M1-3/ Isolated right M/1 Left mandible with M/1-3 in connection with left maxilla with M1-3/ Left mandible + M/1-2 Left mandible + M/1 Left mandible + M/1-3 Left maxilla + M/1-2 Left mandible + M/1-2 Left mandible + M/2-3 Right mandible + M/1-2 Right edentulous mandible, right maxilla + M1/ (continued)
52
C. Denys
Appendix 2.3 (continued) EP number Locality
Level
Anatomical element
1251/01 2433/03 2434/03 2922/00 2923/00 549/01 550/01 552/01 641/01 1066/03 882/03 883/03 990/03 296/05 699/05 700/05 4329/00 1326/03 1327/03 1612/03 191/05 160/03
9S 9S 9S 10E 10E 10E 10E 10E 10 10W 10E 10E 10 10 10W 10W 11 11 11 15 16 17
ULB ULB ULB ULB ULB ULB ULB ULB ULB ULB ULB ULB ULB ULB ULB ULB ULB ULB ULB ULB ULB ULB
1344/05
22E
ULB
Left mandible + M/2-3 Right mandible + M/1-2 Right mandible+/1-3 Left mandible + M/1-2 Right maxilla + M1/ Right mandible + M/1-M/3 Right mandible + M/2 Right mandible + M/2 Right mandible + M/1 Right mandible + M/1-3 Right M1/ Right mandible + M/1 Right mandible + M/1-2 Right maxilla + M1-2/ Left mandible + M/1-2 Left maxilla + M1-2/ Broken M1/ Maxilla + M1/ Left mandible + M/1-2 Right mandible + M/1-2 Cranium Right maxilla + M1-3/ + edentulous mandible maxilla fragment Left mandible + M/1-3
Appendix 2.4 List of specimens attributed to Heterocephalus quenstedti from the Upper Laetolil Beds (ULB) EP number Locality Level Anatomical element 1782/00
2
ULB
1990/00
5
ULB
396/03 781/05 2205/03
5 5 7
ULB ULB ULB
1171/00 326/00
8 8
ULB ULB
4151/00 043/01
8 8
ULB ULB
1427/03 140/05 1059/98 2436/03 258/98 2921/00 3119/00 638/01 639/01 640/01 1067/03 1068/03
8 8 9S 9S 10E 10E 10 10 10 10 10W 10W
ULB ULB ULB ULB ULB ULB ULB ULB ULB ULB ULB ULB
Right + left edentulous maxillae Half cranium with upper molars Right mandible + M/1-2 Right mandible + M/2 Anterior cranial fragment + left M/1-2 + right mandible + M/2 Mandible fragments + incisors Anterior skull fragment + left M/1-2 Left mandible + M/1-3 Right mandible + M/2-3 & anterior cranium with left M1-3/ Left mandible + M/1-2 Right + left mandible + M/1-3 Left mandible + M/1-3 Left mandible + M/1-3 Edentulous left mandible Right mandible + M/1-3 Left mandible + M/2-3 Maxilla fragment + M/1-2 Left mandible + M/1-3 Left mandible + M/1 Right mandible + M/2 Right mandible + M/2-3
1069/03 1082/03 1083/03
10W 10W 10W
ULB ULB ULB
1084/03
10W
ULB
989/03
10
ULB
701/05 883/05 884/05 1784/03 1222/05
10W 10E 10E 22 22E
ULB ULB ULB ULB ULB
Mandible + M/2 Right mandible with M/1-3 Anterior cranial fragment + edentulous left and right maxilla Anterior cranial fragment + left M1/ Associated mandibles left + right + M/1-3 Edentulous left mandible Maxilla fragment Left mandible + M/2-3 Left mandible + M/2-3 Left mandible + M/1
References Black, C. G., & Krishtalka, L. (1986). Rodents, bats, and insectivores from the Plio-Pleistocene sediments to the east of Lake Turkana, Kenya. Contributions in Science, 372, 1–15. Broom, R. (1934). On the fossil remains associated with Australopithecus africanus. South African Journal of Science, 31, 471–480. Broom, R. (1939). The fossil rodents of the limestone cave at Taungs. Annals of the Transvaal Museum, 19, 315–317. Davies, C. (1987). Fossil Pedetidae (Rodentia) from Laetoli. In M. D. Leakey & J. M. Harris (Eds.), Laetoli: A Pliocene site in northern Tanzania (pp. 171–193). Oxford: Clarendon. Denys, C. (1985). Paleoenvironmental and paleobiogeographical significance of the fossil rodent assemblages of Laetoli (Pliocene, Tanzania). Palaeogeography, Palaeoclimatology, Palaeoecology, 52, 77–97. Denys, C. (1987a). Fossil rodents (other than Pedetidae) from Laetoli. In M. D. Leakey & J. M. Harris (Eds.), Laetoli: A Pliocene site in northern Tanzania (pp. 118–170). Oxford: Clarendon. Denys, C. (1987b). Micromammals from the West Natron Pleistocene deposits (Tanzania). Biostratigraphy and paleoecology. Sciences Géologiques Bulletin, 40, 185–201. Denys, C. (1989a). Two new gerbillids (Rodentia, Mammalia) from Olduvai Bed I (Pleistocene, Tanzania). Neues Jahrbuch für Geologie und Paläontologie Abhandlungen, 178, 243–265. Denys, C. (1989b). A new species of bathyergid rodent from Olduvai Bed I (Tanzania, Lower Pleistocene). Neues Jahrbuch für Geologie und Paläontologie Monatshefte, 5, 257–264. Denys, C. (1990). Deux nouvelles espèces d’Aethomys (Rodentia, Muridae) à Langebaanweg (Pliocène, Afrique du Sud): Implications phylogénétiques et paléoécologiques. Annales de Paleontologie, 76, 41–69. Denys, C. (1992). Présence de Saccostomus (Rodentia, Mammalia) à Olduvai Bed I (Tanzanie, Pléistocène inférieur). Implications phylétiques et paléobiogéographiques. Geobios, 25, 145–154. Denys, C. (1999). Of mice and men. Evolution in East and South Africa during Plio-Pleistocene times. In T. Bromage & F. Schrenk (Eds.), African biogeography, climate change and human evolution (pp. 226–252). Oxford: Oxford University Press. Denys, C., & Jaeger, J. J. (1992). Rodents of the Miocene site of Fort Ternan (Kenya) first part: Phiomyids, bathyergids, sciurids and anomalurids. Neues Jahrbuch für Geologie und Paläontologie Abhandlungen, 185, 63–84. Denys, C., Viriot, L., Daams, R., Pelaez-Campomanes, P., Vignaud, P., Andossa, L., & Brunet, M. (2003). A new Pliocene xerine sciurid (Rodentia) from Kossom Bougoudi, Chad. Journal of Vertebrate Palaeontology, 23, 676–687.
2 Rodents Dietrich, W. O. (1942). Altestquartäre Säugetiere aus der südlichen Serengeti, Deutsch-Ostafrika. Palaeontographica, 94A, 43–133. Gentry, A. W. (1987). Pliocene Bovidae from Laetoli. In M. D. Leakey & J. M. Harris (Eds.), Laetoli: A Pliocene site in northern Tanzania (pp. 378–408). Oxford: Clarendon. Geraads, D. (2001). Rongeurs du Miocène supérieur de Chorora, Ethiopie: Murinae, Dendromurinae et conclusions. Paleovertebrata, 30, 89–109. Greenwood, M. (1955). Fossil Hystricoidea from the Makapan Valley, Transvaal. Palaeontologia Africana, 3, 77–85. Greenwood, M. (1958). Fossil Hystricoidea from the Makapan Valley. Transvaal: Hystrix makapanensis nom. nov. for Hystrix major Greenwood. Annals and Magazine of Natural History, 13, 365. Harris, J. M. (1987). Summary. In M. D. Leakey & J. M. Harris (Eds.), Laetoli: A Pliocene site in northern Tanzania (pp. 524–531). Oxford: Clarendon. Hlusko, L. (2007). Earliest evidence for Atherurus and Xenohystrix (Hystricidae, Rodentia) in Africa from the Late Miocene site of Lemudong’o, Kenya. Kirtlandia, 56, 86–91. Jaeger, J. J. (1976). Les Rongeurs (Mammalia, Rodentia) du Pléistocène inférieur d’Olduvai Bed I (Tanzanie). Iè partie: les Muridés. In R. J. G. Savage & S. C. Coryndon (Eds.), Fossil vertebrates of Africa (pp. 58–120). London: Academic. Kingdon, J. (1974). East African mammals, volume II part B (hares and rodents). New York: Academic. Lecompte, E., Aplin, K., Denys, C., Catzeflis, F. M., Chades, M., & Chevret, P. (2008). Phylogeny and biogeography of African Murinae based on mitochondrial and nuclear gene sequences, with a new tribal classification of the subfamily. BMC Evolutionary Biology, 8, 199. http://www.biomedcentral.com/1471-2148/8/199. Manthi, F. K. (2006). The Pliocene micromammalian fauna from Kanapoi, northwestern Kenya, and its contribution to understanding the environment of Australopithecus anamensis. Ph.D. dissertation, University of Cape Town, Cape Town. Manthi, F. K. (2007). A preliminary review of the rodent fauna from Lemudong’o, southwestern Kenya, and its implication to the Late Miocene paleoenvironments. Kirtlandia, 56, 92–105. Maslin, M. A., & Christensen, B. (2007). Tectonics, orbital forcing, global climate change, and human evolution in Africa: Introduction to the African paleoclimate special volume. Journal of Human Evolution, 53, 443–464. Meester, J. A. J., Rautenbach, I. L., Dippenaar, N. J., & Baker, C. M. (1986). Classification of southern African mammals. Transvaal Museum Monographs, 5, 1–359. Mein, P., & Pickford, M. (2006). Late Miocene micromammals from the Lukeino Formation (6.1 to 5.8 Ma), Kenya. Bulletin mensuel de la Société linnéenne de Lyon, 75, 183–223. Mein, P., Pickford, M., & Senut, B. (2000a). Late Miocene micromammals from the Harasib karst deposits, Namibia. Part 1 – Large muroids and non-muroids rodents. Communications of the Geological Survey of Namibia, 12, 375–390. Mein, P., Pickford, M., & Senut, B. (2000b). Late Miocene micromammals from the Harasib karst deposits, Namibia. Part 2A –
53 Myocricetodontinae, Petromyscinae and Namibimyinae (Rodentia, Gerbillinae). Communications of the Geological Survey of Namibia, 12, 391–401. Mein, P., Pickford, M., & Senut, B. (2004). Late Miocene micromammals from the Harasib karst deposits, Namibia. Part 2B – Cricetomyidae, Dendromuridae and Muridae, with a complement on the Myocricetodontonae. Communications of the Geological Survey of Namibia, 13, 43–61. Pickford, M., Mein, P., & Senut, B. (1994). Fossiliferous Neogene karst filling in Angola, Botswana and Namibia. South African Journal of Sciences, 90, 227–230. Pocock, T. N. (1987). Plio-Pleistocene fossil mammalian microfauna of Southern Africa. A preliminary report including description of two new fossil muroid genera (Mammalia, Rodentia). Palaeontologia Africana, 26(7), 69–91. Sabatier, M. (1982). Les rongeurs du site Pliocène à hominidés de Hadar (Ethiopie). Paleovertebrata, 12(1), 1–56. Sénégas, F. (2004). A new species of Petromus (Rodentia, Hystricognatha, Petromuridae) from the early Pliocene of South Africa and its palaeoenvironmental implications. Journal of Vertebrate Paleontology, 24, 757–763. Stehlin, H. G., & Schaub, S. (1951). Die trigonodontie der Simplicidentaten Nager. Schweizerische Paläontologische Abhandlungen, 67, 1–384. Wesselman, H. B. (1984). The Omo micromammals. In M. K. Hecht & F. S. Szalay (Eds.), Contributions to vertebrate evolution. Basel: Karger. Wesselman, H. B., Black, M. T., & Asnake, M. (2009). Small mammals. In Y. Haile-Selassie & G. WoldeGabriel (Eds.), Ardipithecus kaddaba: Late Miocene evidence from the Middle Awash, Ethiopia (pp. 105–134). Berkeley: University of California Press. Wilson, D. E., & Reeder, D. M. (2005). Mammal species of the world. A taxonomic and geographic reference (3rd ed.). Baltimore: The John Hopkins University Press. Winkler, A. J. (1997). Systematic, paleobiogeography and paleoenvironmental significance of rodents from the Ibole Member, Manonga Valley, Tanzania. In T. Harrison (Ed.), Neogene paleontology of the Manonga Valley, Tanzania. (pp. 311–332). New York: Plenum. Winkler, A. J. (2002). Neogene paleobiogeography and East African paleoenvironments: Contributions from the Tugen Hills rodents and lagomorphs. Journal of Human Evolution, 42, 237–256. Winkler, A. J. (2003). Lagomorpha and Rodentia. In M. Leakey & J. M. Harris (Eds.), Lothagam: The dawn of humanity in eastern Africa (pp. 169–198). New York: Columbia University Press. Winkler, A. J., Denys, C., & Avery, M. (2010). Rodentia. In L. Werdelin & W. J. Sanders (Eds.), Cenozoic mammals of Africa (pp. 263–304). Berkeley: California University Press. Woods, C. A., & Kilpatrick, C. W. (2005). Infraorder Hystricognathi. In D. E. Wilson & D. M. Reeder (Eds.), Mammal species of the world, a taxonomic and geographic reference (pp. 1538–1600). Baltimore: The John Hopkins University Press.
wwwwwwwwwwwwwww
Chapter 3
The Lower Third Premolar of Serengetilagus praecapensis (Mammalia: Lagomorpha: Leporidae) from Laetoli, Tanzania Alisa J. Winkler and Yukimitsu Tomida
Abstract The present study suggests evolutionary changes in the morphology and size of the lower third premolar of the leporid Serengetilagus praecapensis from the Upper Ndolanya and Upper Laetolil Beds, Laetoli, Tanzania (ca. 3.85–2.66 Ma). Mandibular depth at p3 was compared also as a proxy indicator of size. The occlusal morphology of p3s from Laetoli is variable, but most commonly the tooth is crescentic with a posteroexternal reentrant (PER) extending about half way across the width of the tooth, plus distinct anteroexternal (AER) and anterior (AR) reentrants. An anterointernal reentrant (AIR) is weak to distinct. A proportionally higher percentage of p3s from the Upper Ndolanya Beds (50%) and the uppermost Upper Laetolil Beds (ULB, between Tuff 7 and the Yellow Marker Tuff, 49%) had a weak AIR compared to only 29% of specimens from between Tuffs 5–7, ULB. The higher frequency of a weak AIR in the geologically younger population is interpreted as the character state being newly reversed to the plesiomorphic condition (AIR weak to absent). There are only two poorly preserved p3s from the Lower Laetolil Beds: on both specimens the AIR and AR are weak to absent (plesiomorphic condition). AR is almost always present on p3s from the Upper Ndolanya and Upper Laetolil Beds. On average, p3s from the Upper Ndolanya Beds are slightly shorter and narrower, and the mandibles slightly less deep at the level of p3 than those from the Upper Laetolil Beds. However, the range of variation of measurements is quite similar between samples from the Upper Ndolanya and Upper Laetolil Beds. A specimen from the Upper Ndolanya Beds (EP 1223/03.1) has a p3 proportionally wider than
A.J. Winkler (*) Roy M. Huffington Department of Earth Sciences, Southern Methodist University, Dallas, TX 75275, USA and Department of Cell Biology, University of Texas Southwestern Medical Center at Dallas, Dallas, TX 75390, USA e-mail:
[email protected] Y. Tomida National Museum of Nature and Science, 3-23-1 Hyakunincho, Shinjukuku, Tokyo 169-0073, Japan e-mail:
[email protected] mean values for other specimens from both the Upper Ndolanya and Upper Laetolil Beds. In conjunction with p3 occlusal morphology, this specimen may represent a new, as yet unnamed, species. Although interesting, the differences observed between samples from the Upper Ndolanya Beds and subunits of the Upper Laetolil Beds are not considered adequate for separation into a distinct species or subspecies. Keywords Pliocene • Rabbit • Phylogeny • Taxonomy
Introduction Laetoli is located on the Eyasi Plateau, in the southern part of the East African Rift, in northern Tanzania. It is one of the most important and prolific paleontological and paleoanthropological localities in Africa, having yielded one of the largest collections of specimens attributable to the early hominin Australopithecus afarensis, as well as a spectacular collection of plants, ichnofossils, invertebrates, reptiles, birds, and other mammals. Laetoli was first collected in the 1930s and was more extensively investigated from 1974–1982 by Mary Leakey and colleagues (Leakey and Harris 1987). Renewed collecting from the area by the Eyasi Plateau Paleontological Expedition under the direction of Terry Harrison from 1998– 2005 has yielded 15,019 mammalian specimens from 60 localities and sub-localities, of which about 34% are lagomorphs (Harrison 2011; Harrison and Kweka 2011). Leakey et al. (1976) noted that lagomorphs were one of the most common taxa from Laetoli, and Leakey (1987: table 1.5) reported that they constituted about 31% of the better-represented taxa from the (Upper) Laetolil Beds and about 5% of the fauna from the Upper Ndolanya Beds. Among the recent collections, lagomorphs represent 38.0% of mammals from the Upper Laetolil Beds, 6.0% from the Lower Laetolil Beds, and 17.4% from the Upper Ndolanya Beds (Harrison 2011). Lagomorph remains from Laetoli (all attributed to Serengetilagus) are primarily isolated incomplete dentitions. Incomplete postcrania are also relatively common and partial skeletons (sometimes articulated) are known, but less common.
T. Harrison (ed.), Paleontology and Geology of Laetoli: Human Evolution in Context. Volume 2: Fossil Hominins and the Associated Fauna, Vertebrate Paleobiology and Paleoanthropology, DOI 10.1007/978-90-481-9962-4_3, © Springer Science+Business Media B.V. 2011
55
56
Overall, the fossil record of African lagomorphs is relatively sparse. It includes ochotonids (minimum of two genera) from the early to middle Miocene of northern, eastern, and southern Africa. Prolagids (one genus) are known only from the late Miocene to early Pleistocene of northern Africa. Fossil leporids (five genera total) are first reported from the late Miocene of eastern Africa. The specimens of Serengetilagus from Laetoli are the most numerous and complete of any known fossil lagomorph from Africa. Serengetilagus praecapensis was described originally by Dietrich (1941, 1942) based on approximately 76 specimens (cranial and postcranial remains: some of this material likely associated with particular individuals) collected by KohlLarsen in 1938–1939 (housed at the Museum für Naturkunde in Berlin). Unfortunately, Dietrich did not designate a holotype. MacInnes (1953) described 21 specimens of Serengetilagus collected by Louis and Mary Leakey in 1935 and housed in the Natural History Museum, London. Although this material is described as coming from Laetoli, Davies (1987: 190) considered it unlikely. Erbaeva and Angermann (1983) provided additional descriptions of Dietrich’s cranial material and designated a lectotype. They noted that the p3 of this species showed great morphologic variability and recognized nine morphotypes (Erbaeva and Angermann 1983: figs. 3, 4). Davies (1987) did a preliminary study of the cranial and postcranial remains (number of examined specimens not indicated) of S. praecapensis collected from Laetoli (Upper Ndolanya and [Upper] Laetolil Beds) by Mary Leakey and now housed at the National Museum of Tanzania, Dar es Salaam, Tanzania. Although remains of Serengetilagus from Laetoli had traditionally been considered to represent a single species, S. praecapensis, Davies (1987) suggested two (unnamed) subspecies were present: one in the Upper Ndolanya Beds and a second in the Upper Laetolil Beds. These subspecies were distinguished primarily on three characters of the auditory region of the cranium. In specimens from the Upper Ndolanya Beds: (1) the auditory bulla is slightly smaller; (2) there is a better developed mastoid flange; and, in particular, (3) the squamosal process above the bulla extends farther posteriorly. Serengetilagus praecapensis has been reported from a few other localities in eastern and central Africa. Four specimens of S. praecapensis are illustrated and briefly described from the Adu-Asa Formation (5.8–5.2 Ma), Middle Awash, Ethiopia (Wesselman et al. 2008). Winkler (2003) described a mandible of S. praecapensis from the early Pliocene (4.22– 4.20 Ma) at Lothagam, Kenya. Serengetilagus aff. S. prae capensis is reported from Kossom Bougoudi, northern Chad (ca. 5 Ma; Brunet et al. 2000). The latter material is noted (without description or illustration) to be primitive with respect to S. praecapensis from Laetoli (Brunet et al. 2000). The only other species of Serengetilagus is S. tchadensis from Toros Menalla, Chad (late Miocene; López-Martínez
A.J. Winkler and Y. Tomida
et al. 2007). This species is known from 18 numbered specimens, some including associated elements, and some probably referable to particular individuals. Cranial and limited postcranial remains are present. Serengetilagus tchaden sis is considered to have some of the more primitive character states of the genus, such as a simpler p3 with only two main external reentrants and upper cheek teeth strongly widened transversely with wear (López-Martínez et al. 2007). Serengetilagus sp. has been noted from several other African localities. It is reported from a minimum of 17 specimens from the late Miocene Lukeino Formation, Tugen Hills, Kenya (Mein and Pickford 2006). Assignment of all this material to Serengetilagus, however, is not considered definitive based on the descriptions given (there are no illustrations) (see discussion in Winkler and Avery 2010). Flynn and Bernor (1987) suggested that an isolated leporid p3 (KNM-KW 138) from the Pliocene Kanam West locality, Kenya, was likely referable to Serengetilagus. Leakey (1965) listed, and very briefly discussed (only a distal tibia was illustrated), lagomorphs from the 1951–1961 excavations at Olduvai Gorge, Tanzania (early Pleistocene). Lagomorphs are reported only from Bed I, and are described as uncommon. They include a large taxon, considered to pertain to Lepus, plus another taxon similar to Serengetilagus (Leakey 1965). Leakey (1971) listed Serengetilagus sp. from the 1960–1963 excavations in Bed I, but there was no discussion or illustration of the material. Lagomorphs of the “Serengetilagus-Trischizolagus group” are reported from Ahl al Oughlam, Morocco (ca. 2.5 Ma; Geraads 2006). Serengetilagus raynali from Grotte des Rhinocéros, Oulad Hamida I, Morocco (Geraads 1994) pertains to Trischizolagus according to several authors (see discussion in Winkler and Avery 2010). In southern Africa, Serengetilagus is listed as occurring in the Plio-Pleistocene of southern Angola (Pickford et al. 1992), although these authors cast some doubt on the identification (Pickford et al. 1992: 20). As noted above, Davies (1987) suggested that more than one taxon of leporid might be present at Laetoli. Erbaeva (personal communication) has also made this suggestion based on p3 morphology. Harrison (personal communication) observed that some remains from the Upper Ndolanya Beds and Upper Laetolil Beds above Tuff 7 were much larger than others from the Upper Ndolanya and Upper Laetolil Beds. Davies (1987) thought that specimens from the Upper Ndolanya Beds were different enough in cranial morphology from those of the Upper Laetolil Beds to warrant subspecies recognition, but he did not mention any differences in p3 morphology or size of the specimens from the two units. The specimens studied by Dietrich (1941, 1942) and MacInnes (1953) were not collected with specific stratigraphic provenance. Davies (1987) even questioned if the material studied by MacInnes (1953) was from Laetoli. Thus, individual specimens from these collections could
3 Lower Third Premolar of Serengetilagus
potentially date from ca. 2.66 to >4.4 Ma (see Stratigraphic Context). Erbaeva and Angermann (1983) noted tremendous variability in p3 morphology in the Dietrich collection, but without stratigraphic control, they could not evaluate if the variability may have been from differences in geologic age. The collection made by Harrison is extensive and has excellent stratigraphic control. Thus, the main goal of the present study is to compare attributes of the p3 (the most significant tooth for morphological comparisons) among the three main stratigraphic horizons (and sub-horizons of the Upper Laetolil Beds) to determine if there is significant variability among the units. The ultimate goals are to better understand the evolutionary history of S. praecapensis and determine if more than one taxon was present at Laetoli.
Stratigraphic Context The Laetoli fossils are derived from several localities within the ‘main Laetoli area’ and from other localities on the Eyasi Plateau. Pliocene deposits from Laetoli are divided into two
57
main units, the Laetolil Beds (³4.4–3.63 Ma) and the Upper Ndolanya Beds (~2.66 Ma). Together, these units crop out over more than 1,600 km2. Leporids are known from the Lower (>4.4–3.85 Ma) and Upper (3.85–3.63 Ma) Laetolil Beds, and the Upper Ndolanya Beds (2.66 Ma). Leporids have not been reported (Leakey or Harrison collections) from younger units in the area, the Olpiro and Ngaloba Beds. All specimens were surface collected, but in most cases derived from known stratigraphic horizons. A stratigraphic profile for Laetoli is illustrated in Fig. 3.1. The Laetoli sediments include aeolian, airfall, and waterlain tuffs that are interbedded with mafic lavas and medium- to fine-grained epiclastics (Deino 2011). In stratigraphic thickness, the Upper Ndolanya Beds are from about 8–16 m thick (Ditchfield and Harrison 2011) and consist primarily of clayey aeolian tuffs (Hay 1987). The Upper Laetolil Beds are about 44–59 m in thickness: the Lower Laetolil Beds, where well developed, may be 64 m in thickness (Hay 1987). The Upper Laetolil Beds are primarily composed of aeolian tuff, but contain some airfall tuffs (Hay 1987). The Lower Laetolil Beds are composed largely of aeolian tuff interbedded with airfall and waterlain tuffs (Hay 1987). Lagomorphs examined
Fig. 3.1 Stratigraphic profile for Laetoli, Tanzania (From Harrison and Kweka (2011); based on Drake and Curtis (1987); Hay (1987); Ndessokia (1990); Manega (1993); Ditchfield and Harrison (2011); Deino (2011))
58
for this study from the Upper Laetolil Beds had been collected with reference to eight marker tuffs, with Tuff 1 the oldest and Tuff 8 the youngest (Fig. 3.1). The Upper Ndolanya Beds are well sampled and have yielded an extensive fauna, of which about 17% of all specimens of mammals are lagomorphs (Harrison collection; Leakey 1987: table 1.5 reported about 5%). However, there is a strong taphonomic bias against small mammals in this unit, which may be reflected in the lower proportion of lagomorphs compared to the Upper Laetolil Beds (Harrison, personal communication). The Lower Laetolil Beds have not yet been sampled adequately (i.e., the collection from these beds is about 1% of the size of that from the upper unit): of the 258 specimens collected, about 6% are lagomorphs (Harrison 2011). Small mammals comprise a large proportion of the wellsampled fauna from the Upper Laetolil Beds. In the Harrison collection, about 37% of all specimens of mammals from the Upper Laetolil Beds are lagomorphs (Harrison 2011). In the Leakey collection, lagomorphs represent about 31% of the mammals from the Upper Laetolil Beds (Leakey 1987: table 1.5).
Materials and Methods Lagomorph crania constitute the most useful element for identification. However, even relatively complete crania are extremely rare from Laetoli, and most cranial remains are incomplete mandibles and maxillae. Of this material, the most diagnostic tooth is the p3. Hence, the present study focuses on p3 morphology and occlusal size (i.e., length, width, length/width, depth of posteroexternal reentrant, depth of posteroexternal reentrant/width of tooth). When possible, the height of the mandible at p3 was also measured as an additional indicator of size of the specimen. Measurements of occlusal and alveolar toothrow length would also have been useful adjunct indicators of size of the specimens, but too few toothrows were preserved for these measurements to be helpful. All p3s from the Upper Ndolanya Beds (N = 40) and Lower Laetolil Beds (N = 2) were examined. About half (N = 168) of the p3s from the Upper Laetolil Beds were included. Specimens from the Upper Ndolanya Beds are from five localities (7E, 15, 18, 22S, and Silal Artum). Specimens from the Lower Laetolil Beds are from Kakesio 8. Specimens from the Upper Laetolil Beds include: 64 from between Tuffs 7 and the Yellow Marker Tuff (Locs. 1, 3, 7, 8, 11, 15–17), 76 from between Tuffs 5 and 7 (Locs. 2, 8, and 10E), 13 from between Tuffs 3 and 5 (Loc. 5), and 15 from below Tuff 3 (Locs. 9S, 10, and 10W). Specimens studied (210 total from all units) are listed in the Appendix 3.1.
A.J. Winkler and Y. Tomida
Fig. 3.2 Tooth terminology for leporid p3. AER, anteroexternal reentrant (= protoflexid); AIR, anterointernal reentrant (= paraflexid); AR, anterior reentrant (= anteroflexid); PER, posteroexternal reentrant (= hypoflexid); PIR, posterointernal reentrant (= mesoflexid) (Tooth terminology modified from White 1991)
Tooth terminology for a leporid p3 (from White 1991) is illustrated in Fig. 3.2. Measurements of the occlusal surface of the teeth were made using a reticule in a Wild Heerbrugg dissecting stereomicroscope. In addition to measurements of the p3, the depth of the mandible was measured at the level of p3 using Mitutoyo digital calipers. Measurements were taken only on specimens believed to be adults. Specimens were considered juveniles (and excluded) if they had any of the following characteristics: cone-shaped p3, dp3 or dp4 present, or they were of proportionally smaller size (often including a more gracile mandible and more porous appearing, less calcified bone) than other specimens. Morphology of juvenile p3s was, however, noted. The pattern of enamel reentrants on p3 is of taxonomic significance for leporids, and is the most widely used criterion for classification. Studies of modern and fossil populations have demonstrated, however, that although the vast majority of individuals within a population possess the diagnostic p3 pattern (or slight variants of it), there are often individuals with patterns that might be considered diagnostic of other taxa (Hibbard 1963). Sometimes these aberrant patterns are observed more commonly in younger individuals: certain reentrants may not persist along the full length of the crown, so may be “present” or “not present” depending upon occlusal wear. For this reason, one needs to look carefully at morphological variation in a fossil sample to decide if one or multiple taxa are represented. Erbaeva and Angermann (1983) classified 143 p3s of S. praecapensis from the Dietrich collection into nine morphotypes based on the presence or absence of particular reentrants plus the development of some of these reentrants (I–IX; Fig. 3.3). Their classification scheme included: I, only PER and AER present; II, PER, AER, and AR present; III, PER, AER, and AR (doubled or branched) present; IV, PER, AER, AR, and AIR present; V, PER, AER, AR (doubled or branched), and AIR present; VI, PER, AER, AR, AIR, and large PIR present; VII, PER, AER, AR, AIR, and small PIR present; VIII, PER, AER, and AIR present, but AR absent;
3 Lower Third Premolar of Serengetilagus
Fig. 3.3 Camera lucida drawings of Serengetilagus praecapensis p3s illustrating the nine occlusal morphotypes of Erbaeva and Angermann (1983: I–IX). Illustrations are the same teeth figured by Erbaeva and Angermann (1983: fig. 3), but the original teeth were redrawn to show enhanced detail of the enamel pattern and thickness. (I) (Winkler and Tomida, morphotype A), MB.Ma 1449/4 (only PER and AER present); (II) (type B), MB.Ma 1451/46 (PER, AER, AR); (III) (type B), MB.Ma 1450/2 (PER, AER, AR with double or multiple crenulations); (IV) (type B+), Mb.Ma 1447/6 (PER, AER, AR, AIR); (V) (type B+), MB.Ma 1449/3 (PER, AER, AIR, AR with double or multiple crenulations); (VI) (type C), MB.Ma 1448/3 (PER, AER, AR, AIR, PIR large); (VII) (type C), MB.Ma 1450/15 (PER, AER, AR, AIR, PIR small); (VIII) (no corresponding type), MB.Ma 1447/17 (PER, AER, AIR present, AR absent); (IX) (type C), MB.Ma 1451/44 (PER, AER, AR, AIR, enamel lake)
IX, PER, AER, AR, AIR, and enamel lake present. The most common morphotype seen in their sample was IV (PER, AER, AR, and AIR present; 57% of the sample). For the present study, we noted a wide range of variation in the size of AIR and found it was often difficult to classify objectively extremely small AIRs as present or absent. Often the AIR would be present as a slight indentation on the occlusal surface (e.g., Fig. 3.3 II; which would have been classified as absent by Erbaeva and Angermann 1983), but it was observed still as a distinct groove on the side of the tooth. Thus, we used a modified classification scheme to more precisely incorporate the variability of AIR. We used five morphotypes: A, only PER and AER present (= morphotype I of Erbaeva and Angermann 1983); B, AER, PER, AR (single or multiple crenulations) present, and AIR completely absent (= morphotypes II and III); B-, AER, PER, AR (single or multiple crenulations) present, and AIR very weak
59
Fig. 3.4 Camera lucida drawings of representative examples of Serengetilagus praecapensis p3 occlusal morphology from the Harrison collections. The specimens are from (a) the Lower Laetolil Beds, (b–d) Upper Laetolil Beds, and (e–h) Upper Ndolanya Beds. (a) EP 208/03, morphotype B; (b) EP 902/03.1, type B+; (c) EP 554/01.2, type B+; (d) EP 1565/98.1, type B−; (e) EP 1223/03.1, type B+; (f) EP 810/01.1, type B+; (g) EP 3475/00.1, type B−; (h) EP 810/01.2, type A
(may be observed only as a weak indentation on the occlusal surface, but present as a groove on the side of the tooth; Erbaeva and Angermann 1983, did not classify this morphotype); B+, AER, PER, AR (single or multiple crenulations), and distinct AIR present (= morphotypes IV and V); C, AER, PER, AR, AIR, and PIR (relatively large or small or as an enamel lake) present (= morphotypes VI, VII, IX). No specimens had a distinct AIR, but lacked AR and PIR (morphotype VIII). A very weak groove in the area where PIR would be located was sometimes observed along the side of the tooth: sometimes this groove was expressed as a slight indentation on the occlusal surface. These very weak grooves or indentations were not considered distinct enough to count PIR as present. Specimens illustrated in Fig. 3.3 and representative specimens of Serengetilagus from the Harrison collection (Fig. 3.4) were made by YT using a camera lucida. Final illustrations were made by tracing these pencil drawings using Adobe Illustrator CS4. Statistical analyses were made using SPSS 13.0. Institutional abbreviations. KNM-KW, National Museums of Kenya, Kanam West locality; KNM-LT, National Museums of Kenya, Lothagam locality; M, Natural History Museum, London; MB.Ma, Museum für Naturkunde, Berlin.
60
A.J. Winkler and Y. Tomida
Anatomical abbreviations for p3. AER, anteroexternal reentrant (= protoflexid); AIR, anterointernal reentrant (= paraflexid); AR, anterior reentrant (= anteroflexid); PER, posteroexternal reentrant (= hypoflexid); PIR, posterointernal reentrant (= mesoflexid). Tooth terminology modified from White (1991).
has a short PER extending about half-way across the occlusal surface, plus an AR, AER, and an AIR; rudimentary PIR extremely rare; mean p3 length is 3.3 mm and mean width is 3.1 mm; occlusal shape of p3 variable, but mostly anteroposteriorly elongated.
Occlusal Morphology
Results Dietrich (1941, 1942: 58); translation from López-Martínez et al. 2007) diagnosed S. praecapensis, stating that on p3 the PIR was either extremely reduced, or, in most cases, absent, and the other reentrants were as usually seen in Palaeolaginae (i.e., presence of PER, AER, AR, AIR). López-Martínez et al. (2007: 4) provided a generic diagnosis and reiterated the “Extended Diagnosis” of S. praecapensis by Erbaeva and Angermann (1983: 59). López-Martínez et al. (2007) noted that Serengetilagus belongs in Archaeolaginae, not Palaeolaginae, because the former subfamily includes taxa with the PIR usually absent. For p3, the generic diagnosis (López-Martínez et al. 2007: 4) includes “p3 crescentic in shape with two main, constant external folds [PER (hypoflexid) extending about half-way across the crown and shallow AER (protoflexid)] and up to three additional folds variably present [AR (anteroflexid) variably developed, weak AIR (paraflexid) and exceptionally a PIR-enamel lake (mesoflexid-mesofossetid), mainly in young individuals]; when an AR (anteroflexid) is present, lingual anteroconid is weaker than the labial one.” The diagnosis for S. praecapensis by Erbaeva and Angermann (1983: 59) includes (for p3): the most common morphotype
Table 3.1 presents the frequency of each p3 morphotype (A, B, B−, B+, C) in the Harrison collection based on stratigraphic position. Figure 3.4 illustrates representative specimens from the Upper Ndolanya Beds and the Upper and Lower Laetolil Beds. The two most common morphotypes are B − and B+: PER, AER, AR, and AIR are present. Morphotype B − (Fig. 3.4d, g) has a weak AIR and B + (Fig. 3.4b, c, e, f) has a distinct AIR. Although the sample sizes from all units are relatively low, it is noteworthy that S. praecapensis from the younger part of the section (Upper Ndolanya Beds, [UNB], and between Tuffs 7 and YMT [Yellow Marker Tuff], Upper Laetolil Beds [ULB]) more commonly exhibits a weak AIR (B−, about 49–50% of specimens) than one that is distinct (B+, 26%). This pattern is reversed in the next geologically older sample, from between ULB Tuffs 5 and 7, where a distinct AIR is more commonly observed (B+, 51%; B−, 29%). Specimens collected below Tuff 5 also more commonly have a distinct AIR, however, sample sizes are very low. The distribution of the frequency of morphotype B, complete absence of AIR, does not show a meaningful pattern: this is likely because there are so few teeth overall with that morphotype. Abundances of all other morphotypes are low. White’s (1991) diagnosis of the most primitive leporid,
Table 3.1 Lower third premolar morphotypes of Serengetilagus praecapensis (adults and juveniles) from Laetoli, Tanzania
Morphotypes
Stratigraphic unit (Harrison collection; total N = 191) Upper Laetolil Beds Upper Ndolanya Between Between Beds T7 and YMT T5 and T7
Between T3 and T5
Below Tuff 3
Dietrich collection
A 1 (2.9%) 0 0 0 0 5 (3.5%) B 3 (8.8%) 6 (10.5%) 2 (2.8%) 2 (15.4%) 1 (6.7%) 31 (21.7%) B− 17 (50.0%) 28 (49.1%) 21 (29.2%) 3 (23.1%) 4 (26.7%) Not applicable B+ 9 (26.5%) 15 (26.3%) 37 (51.4%) 7 (53.8%) 10 (66.7%) 97 (67.8%) C 4 (11.8%) 8 (14.0%) 12 (16.7%) 1 (7.7%) 0 6 (4.2%) VIII 0 0 0 0 0 4 (2.8%) Total number of 34 57 72 13 15 143 specimens (100%) Morphotype designations: A, only PER and AER present (= morphotype I of Erbaeva and Angermann (1983)); B, AER, PER, AR (single or multiple crenulations) present, and AIR absent (= morphotypes II and III); B−, AER, PER, AR (single or multiple crenulations) present, and AIR very weak (may be observed only as a weak indentation on the occlusal surface, but present as a groove on the side of the tooth); B+, AER, PER, AR (single or multiple crenulations), and distinct AIR present (= morphotypes IV and V); C, AER, PER, AR, AIR, and PIR (large or small or as an enamel lake) present (= morphotypes VI, VII, IX). No specimen from the Harrison collection has a distinct AIR, but lacks AR and PIR (morphotype VIII). Values for Dietrich collection from Erbaeva and Angermann (1983) T Tuff, YMT Yellow Marker Tuff
3 Lower Third Premolar of Serengetilagus
Alilepus (here considered the outgroup), stated that AIR is usually absent: thus, absence (or, by implication, weakness) of AIR could be considered the plesiomorphic condition for the Leporidae. Averianov (1999) had also considered the lack of AIR to be the plesiomorphic condition. The lower frequency of p3s with a distinct AIR in the Upper Ndolanya Beds and uppermost part of the Upper Laetolil Beds is interpreted as this character newly reverting to the plesiomorphic condition. The vast majority of the p3s in the Harrison collection (from all stratigraphic units) are from adult individuals: 33 are classified as juveniles, using the criteria in Materials and Methods. The distribution of morphotypes of the juveniles includes A (0%), B (N = 2, 9.1%), B − (N = 9, 27.3%), B + (N = 9, 27.3%), and C (N = 12, 36.4%; characterized by presence of a PIR or enamel lake in the area PIR would be located). One specimen (EP 1629/04) is very young (unworn) and is close to morphotype C, but it lacks AR at the occlusal surface and on the side of the tooth. The total sample of juveniles is too small to analyze by stratigraphic unit. In the younger specimens, PER, AER, AIR and PIR are present at the occlusal surface: PER and PIR are comparable in size. Slightly older specimens have a cement-filled groove for AR further down the crown, which would be exposed with occlusal wear (e.g., EP 810/01.5). Some very young specimens with a large cement-filled PIR at the occlusal surface (e.g., EP 3296/00.1) show this reentrant greatly reduced (and lacking cement) at the base of the crown. Of the total sample of specimens (adults and juveniles from the UNB and Laetolil Beds) with a PIR or enamel lake (morphotype C, N = 25), 48% are juveniles and about 7% are adults. Although our sample of very young individuals is extremely small, it appears that developmentally the distinct PIR observed in the youngest p3s is lost eventually through occlusal wear. The PIRs (or enamel lakes) seen in some adult S. praecapensis represent retention (in part) of the juvenile pattern. The frequencies of morphotypes of specimens examined by Erbaeva and Angermann (1983) from the Dietrich collection are also listed in Table 3.1, with the morphotype designations adjusted to the system used in the present paper. The most common morphotype (B+) is characterized by the presence of a PER, AER, AR, and AIR (weak or distinct). Morphotype B (AIR absent) is observed more commonly in the Dietrich than in the Harrison collection, because some specimens with a weak AIR may have been classified as having AIR absent. Because different classification systems were used, one cannot use the distribution of morphotypes to decipher where the Dietrich material was most likely collected from within the Laetoli stratigraphic section. Only two p3s are known currently from the Lower Laetolil Beds (EP 208/03 and EP 1508/03). Neither of these specimens is preserved sufficiently to measure, but a drawing from the cross-section of EP 208/03 (Fig. 3.4a) illustrates a
61
tooth with morphotype B and comparable in size to other specimens of S. praecapensis. Compared to most other specimens of S. praecapensis, however, the AR of EP 208/03 is shallow and does not appear to extend to the base of the crown. The AER has a distinctive additional crenulation observed uncommonly in S. praecapensis. EP 1508/03 appears to lack AR and PIR, and the AIR is very small (close to morphotype A). As for AIR, an AR that is extremely weak to absent is the plesiomorphic condition (with Alilepus as the outgroup; White 1991; Averianov 1999).
Measurements Occlusal measurements of p3 and measurement of the depth of the mandible at p3 are given in Table 3.2. Unlike morphology of the p3, where the distribution of morphotypes is most similar between the Upper Ndolanya Beds and between Tuff 7-YMT, mean values are most similar between Tuff 7-YMT and Tuff 5–7 (sample sizes from below T5 are very low: all measurements from these units are within the ranges seen for other units). This is observed for p3 length and width, and mandibular depth at p3. For these measurements, the mean values of specimens from the Upper Ndolanya Beds are slightly lower. However, the ranges of values from all three units are comparable, although a couple of exceptions from the Upper Ndolanya Beds should be noted. For p3 width, one specimen (EP 3296/00.1) has a p3 that is narrower (2.58 mm) than any other p3 measured. This specimen is broken, and the measurement was made on the crosssection of the tooth, so this may have introduced some error to the reading. The next narrowest p3 (from any unit) is 2.75 mm. The narrow width of EP 3296/00.1 affects the length to width ratio resulting in the high value of 1.32 for the Upper Ndolanya Beds. The next upper value for p3 length to width (L:W) ratio for the UNB is 1.21, which is closer to the upper values for the other units. The very wide p3 from the Upper Ndolanya Beds (EP 1223/03.1, width = 3.92 mm; the next widest tooth is 3.50 mm) is discussed in more detail below. The p3 length to width ratio, depth of the PER, and depth of the PER to tooth width ratio are comparable among specimens from all units, including the mean values and the observed ranges. Unfortunately, sample sizes from the Upper Ndolanya and Upper Laetolil Beds are relatively small, so statistical comparisons among the units for the different parameters may not be meaningful. Larger sample sizes may confirm that S. praecapensis from the Upper Ndolanya Beds is smaller on average in some measurements, but the observed difference may also be an artifact of sample size. Erbaeva and Angermann (1983) provided measurements of specimens in the Dietrich collection (Table 3.2). The
62
A.J. Winkler and Y. Tomida
Table 3.2 Measurements (mm) of the p3 and mandibular depth at p3 of adult Serengetilagus praecapensis from Laetoli, Tanzania Stratigraphic unit (Harrison collection) Upper Laetolil Beds Upper Between Tuff 7 Between Between Below Lower Measurements Ndolanya Beds and YMT Tuffs 5 and 7 Tuffs 3 and 5 Tuff 3 Laetolil Beds
Dietrich collection
p3 length X SD OR N
3.26 0.18 3.00–3.67 24
3.35 0.21 2.92–3.75 24
3.36 0.16 3.08–3.75 29
3.33 0.08 3.25–3.42 5
3.22 0.14 3.08–3.42 6
– – – –
3.27 0.02 2.9–3.7 54
p3 width X SD OR N
3.05 0.31 2.58–3.92 23
3.26 0.20 2.75–3.50 23
3.26 0.22 2.75–3.58 28
3.19 0.18 2.92–3.33 4
3.25 0.16 3.00–3.42 5
– – – –
3.08 0.03 2.7–3.5 52
p3 length/width X SD OR N
1.07 0.10 0.88–1.32 22
1.03 0.07 0.90–1.16 22
1.03 0.07 0.93–1.13 29
1.04 0.05 1.00–1.11 4
0.98 0.06 0.93–1.06 5
– – – –
1.07 – 0.91–1.21 52
Depth of PER X SD OR N
1.44 0.18 1.08–1.83 23
1.42 0.19 1.08–1.75 17
1.47 0.17 1.17–1.75 25
1.23 0.04 1.17–1.25 4
1.28 0.27 1.00–1.67 5
– – – –
1.38 0.02 1.0–1.8 52
Depth PER/width p3 X SD OR N
0.48 0.05 0.38–0.56 22
0.43 0.06 0.32–0.54 16
0.44 0.06 0.32–0.59 25
0.38 0.03 0.36–0.43 4
0.38 0.06 0.31–0.43 4
– – – –
0.45 – 0.34–0.56 51
10.68 0.67 10.11–11.98 6
11.05 0.61 10.17–11.82 6
– – 11.72–12.02 2
11.11 0.15 9.7–13.0 33
Mandibular depth at p3 X 10.49 11.36 11.16 SD 0.55 0.73 0.64 OR 9.65–11.82 9.47–12.50 9.71–12.27 N 21 14 23 Measurements of Dietrich collection from Erbaeva and Angermann (1983) X mean, SD standard deviation, OR observed range, N number of specimens
ranges of values for this larger sample are quite similar to those for specimens in the Harrison collection from all stratigraphic units in all measurements and ratios (except for the two teeth discussed above). In mean values, however, the measurements for p3 length and width for specimens from the Dietrich collection are closer to those from the Upper Ndolanya Beds than the Upper Laetolil Beds. Mean mandibular depth at p3 for the Dietrich collection is more similar to values from the Upper Laetolil Beds than the smaller value for the Upper Ndolanya Beds. As mentioned earlier regarding differences in mean size in the Harrison collection, comparisons between the Dietrich and Harrison collections are plagued by issues of sample size. It is premature, at this point, to speculate on whether the Dietrich collection samples primarily the Upper Ndolanya or Upper Laetolil Beds (or a subunit within the ULB), or a combination of the two.
As also noted earlier, some dental and postcranial remains of specimens from the Upper Ndolanya Beds appear to be extremely large compared to other specimens from the Upper Ndolanya and Upper Laetolil Beds. Only one specimen from the Upper Ndolanya Beds had a p3 that caught our attention as being unusually large, EP 1223/03.1, an incomplete mandible. On a scatter plot of p3 length versus width (Fig. 3.5), the p3 of EP 1223/03.1 (circle) is distinctly wider than other p3s from the Upper Ndolanya and Upper Laetolil Beds. This tooth is at the upper end of the range of variation for p3 length. The p3 of EP 1223/03.1 is statistically significantly wider than the p3s from both the Upper Ndolanya and the Upper Laetolil Beds at p 1,600 km2), amount of geologic time represented (from older than 4.3 to ~2.66 Ma), and tremendous sample size (>4,000 specimens) from Laetoli, evidence of evolutionary change and even the presence of multiple taxa would not be unexpected. Comparing p3 morphology of the samples from the Upper Ndolanya Beds and sediments between Tuff 7 and YMT, ULB, to specimens from between Tuffs 5 and 7, ULB, a weak AIR (49–50%) is more commonly observed in the
A.J. Winkler and Y. Tomida
upper younger units than in the older units (weak AIR 29%). A weak to absent AIR is likely the plesiomorphic condition in leporids: its lower frequency in the upper units is interpreted as this character newly reversing to the plesiomorphic condition. Only two poorly preserved p3s are currently known from the Lower Laetolil Beds: on both specimens the AIR and AR are weak to absent. AR is almost always present (derived condition) on p3s from the younger horizons. Although sample sizes are small and the observed ranges of values overlap, p3s from the Upper Ndolanya Beds are on average slightly shorter and narrower than those from the Upper Laetolil Beds. The length to width ratio, depth of the PER, and depth of the PER relative to the width of the tooth, are similar in both samples. The mandibles from the Upper Ndolanya Beds tend also to be less deep at the level of p3 compared to those from the Upper Laetolil Beds. A single mandible from the Upper Ndolanya Beds (EP 1223/03.1) has a p3 larger and proportionally wider, and a mandible deeper at the level of p3 than mean values for other specimens from the Upper Ndolanya. In conjunction with its unusual p3 occlusal morphology, this specimen may represent a new, as yet unnamed, species. The results of this study suggest some differences between the p3s of S. praecapensis from the Upper Ndolanya Beds and the subunits of the Upper Laetolil Beds. However, sample sizes are relatively small, and these differences are not considered significant enough to warrant specific (or subspecific) separation. The ranges of measurements and morphology overlap, and specimens, if collected individually, could not be identified confidently as belonging to one population or the other. Laetoli has also produced abundant fragmentary maxillary and postcranial remains: nearly complete crania and associated skeletons are present but rare (crania with mandibles extremely rare). It will be interesting to see if differences in size and morphology suggested by this study are also present in other skeletal elements. Davies (1987) had observed morphological differences in the crania of some specimens from the Upper Ndolanya Beds compared to those from the Upper Laetolil Beds. It will be important to see if his observations can be confirmed, and if differences in cranial morphology are associated with any differences in p3 size or morphology. This should provide a more in depth picture of the evolution of this interesting animal. Acknowledgments We are grateful to T. Harrison for the invitation to study the Laetoli Serengetilagus, for discussions about the Laetoli fauna and stratigraphy, and for providing us with Fig. 3.1. Paul Msemwa, Director, and A. Kweka, Senior Curator (National Museums of Tanzania), Andy Currant (Natural History Museum, London), and W.-D. Heinrich and O. Hampe (Museum für Naturkunde, Berlin, Dietrich collections) kindly provided access to specimens of Serengetilagus. This manuscript benefited from engaging discussions on leporids with M. Erbaeva and R. Angermann. We thank D. Winkler
3 Lower Third Premolar of Serengetilagus and three anonymous reviewers for critiquing a draft of this manuscript. Funding for fieldwork at Laetoli and for specimen research was provided by the National Geographic Society, the Leakey Foundation, and the National Science Foundation (grants BCS-9903434 and BCS-0309513) to T. Harrison. Additional funding for research was provided by the Japan Society for the Promotion of Science (grant No. 18540464) to Y. Tomida.
Appendix 3.1 Specimens of Serengetilagus praecapensis from Laetoli, Tanzania, included in this study Horizon Locality Field numbers Upper Ndolanya Beds – 7E – 15 – – – Upper Laetolil Beds Between T7 and YMT Between T7 and 2 m above T8 Between T7 to just above T8 Between T7 and T8
1480/00.1–.2; 1223/03.1–.4 208/04.1–.2; 1064/01.1–.2; 3296/00.1–.3; 3475/00.1; 3488/00; 4049/00 18 092/03.1–.3; 282/04.1–.3; 810/01.1–.5; 992/00.1–.7; 2355/00.1–.2 22S 1824/03.1–.2 Silal Artum 560/04; 1540/01.1 1 17 16
221/00.1–.3; 1429/00 1629/04; 1633/04; 2326/00 1480/00.1–.2; 1223/03.1–.4
11 16 1 3
1081/04.1–.2 276/03; 599/00.1–.2 512/04; 1897/03.1–.2 478/03.1–.3; 1627/00.1–.7; 2760/00.1–.2 2213/03.1–.3; 2289/01.1 258/00.1–.3; 1492/03; 1493/03A (2 specimens) 1319/03.1–.3; 2586/00.1–.5; 4309/00.1–.2 1448/98.1–.6 932/01.1–.4 996/01.1–.4 070/99.1–.5; 234/98.1–.22; 380/98.1–.6; 1565/98.1 337/01.1–.3; 503/01; 609/03.1–.2 373/04 049/04.1–.3; 080/04.1–.4; 554/01.1–.7; 798/00.1–.2; 902/03.1–.14; 1267/04.1–.3; 2893/00.1–.2 261/01; 425/03.1–.3; 1388/04.1–.2; 1976/00.1–.5; 2806/00.1–.2 629/01 1050/98.1–.2; 1266/01.1–.2; 2444/03.1–.2 521/98.1–.3; 882/98; 182/99.1–.2 739/98.1–.2
7 8 11
Between T6 and T7
15 7 11 10E
Between T5 and T7
2
Above T7
8 10E
Between T3 and T5
5
Below T3 Below T2
10W 9S 10 10W
Lower Laetolil Beds Kakesio 8 208/03 Kakesio 2–4 1508/03 All field numbers begin with prefix EP. Format for field numbers is specimen/year collected. Numbers to right of year (e.g., 1452/00.1) indicate individual specimens grouped under one field number. T Tuff; YMT Yellow Marker Tuff
65
References Averianov, A. O. (1999). Phylogeny and classification of Leporidae (Mammalia, Lagomorpha). Vestnik zoologii, 33, 41–48. Brunet, M., & MPFT (2000). Chad: Discovery of a vertebrate fauna close to the Mio-Pliocene boundary. Journal of Vertebrate Paleontology, 20, 205–209. Davies, C. (1987). Note on the fossil Lagomorpha from Laetoli. In M. D. Leakey & J. M. Harris (Eds.), Laetoli: A Pliocene site in northern Tanzania (pp. 190–193). Oxford: Clarendon. Deino, A. L. (2011). 40Ar/39Ar dating of Laetoli, Tanzania. In T. Harrison (Ed.), Paleontology and geology of Laetoli: Human evolution in context (Geology, geochronology, paleoecology and paleoenvironment, Vol. 1, pp. 77–97). Dordrecht: Springer. Dietrich, W. O. (1941). Die säugetierpaläontologische Ergebnisse der Kohl-Larsen’schen Expedition, 1937–1939 im nordlichen DeutschOstafrika. Centralblatt für Mineralogie, Geologie und Paläontologie, Stuttgart, 1941B, 217–223. Dietrich, W. O. (1942). Altestquartäre Säugetiere aus der Südlichen Serengeti, Deutsch-Ostafrikas. Palaeontographica, Stuttgart, 94(A), 43–133. Ditchfield, P., & Harrison, T. (2011). Sedimentology, lithostratigraphy and depositional history of the Laetoli area. In T. Harrison (Ed.), Paleontology and geology of Laetoli: Human evolution in context (Geology, geochronology, paleoecology, and paleoenvironment, Vol. 1, pp. 47–76). Dordrecht: Springer. Drake, R., & Curtis, G. H. (1987). K–Ar geochronology of the Laetoli fossil localities. In M. D. Leakey & J. M. Harris (Eds.), Laetoli. A Pliocene site in northern Tanzania (pp. 48–52). Oxford: Clarendon. Erbaeva, M. A., & Angermann, R. (1983). Das originalmaterial von Serengetilagus praecapensis Dietrich, 1941 – ergänzende beschreibung und vergleichende diskussion. Schriftenreihe für Geologische Wissenschaften, 19/20, 39–60. Flynn, L. J., & Bernor, R. L. (1987). Late Tertiary mammals from the Mongolian People’s Republic. American Museum Novitates, 2872, 1–16. Geraads, D. (1994). Rongeurs et lagomorphes du Pleistocène moyen de la ‘Grotte des Rhinocéros’, Carrière Oulad Hamida 1 à Casablanca, Maroc. Neues Jahrbuch für Geologie und Paläontologie Abhandlungen, 191, 147–172. Geraads, D. (2006). The late Pliocene locality of Ahl al Oughlam, Morocco: Vertebrate fauna and interpretation. Transactions of the Royal Society of South Africa, 61, 97–101. Harrison, T. (2011) Laetoli revisited: Renewed palaeontological and geological investigations at localities on the Eyasi Plateau in northern Tanzania. In T. Harrison (Ed.), Paleontology and geology of Laetoli: Human evolution in context (Geology, geochronology, paleoecology and paleoenvironment, Vol. 1, pp. 1–15). Dordrecht: Springer. Harrison, T., & Kweka, A. (2011). Paleontological localities on the Eyasi Plateau, including Laetoli. In T. Harrison (Ed.), Paleontology and geology of Laetoli: Human evolution in context (Geology, geochronology, paleoecology and paleoenvironment, Vol. 1, pp. 17–45). Dordrecht: Springer. Hay, R. L. (1987). Geology of the Laetoli area. In M. D. Leakey & J. M. Harris (Eds.), Laetoli: A Pliocene site in northern Tanzania (pp. 23–47). Oxford: Clarendon. Hibbard, C. W. (1963). The origin of the P3 pattern of Sylvilagus, Caprolagus, Oryctolagus and Lepus. Journal of Mammalogy, 44, 1–15. Leakey, L. S. B. (1965). Olduvai Gorge 1951–1961 (Vol. I). Cambridge: Cambridge University Press. Leakey, M. D. (1971). Olduvai Gorge (Excavations in Beds I and II, Vol. 3). Cambridge: Cambridge University Press. Leakey, M. D. (1987). Introduction. In M. D. Leakey & J. M. Harris (Eds.), Laetoli: A Pliocene site in northern Tanzania (pp. 1–22). Oxford: Clarendon. Leakey, M. D., & Harris, J. M. (Eds.). (1987). Laetoli: A Pliocene site in northern Tanzania. Oxford: Oxford University Press.
66 Leakey, M. D., Hay, R. L., Curtis, G. H., Drake, R. E., Jackes, M. K., & White, T. D. (1976). Fossil hominids from the Laetolil Beds. Nature, 262, 460–466. López-Martínez, N., Likius, A., Mackaye, H. T., Vignaud, P., & Brunet, M. (2007). A new lagomorph from the Late Miocene of Chad (Central Africa). Revista Española de Paleontologia, 22, 1–20. MacInnes, D. G. (1953). The Miocene and Pleistocene Lagomorpha of East Africa. Fossil Mammals of Africa, 6, 1–30. Manega, P. C. (1993). Geochronology, geochemistry and isotopic study of the Plio-Pleistocene hominid sites and the Ngorongoro volcanic highland in northern Tanzania. Ph.D. dissertation, University of Colorado, Boulder. Mein, P., & Pickford, M. (2006). Late Miocene micromammals from the Lukeino Formation (6.1 to 5.8 Ma), Kenya. Bulletin et Mémoires de la Société Linnéen de Lyon, 75, 183–223. Ndessokia, P. N. S. (1990). The mammalian fauna and archaeology of the Ndolanya and Olpiro Beds, Laetoli, Tanzania. Ph.D. dissertation, University of California, Berkeley.
A.J. Winkler and Y. Tomida Pickford, M., Mein, P., & Senut, B. (1992). Primate bearing Plio-Pleistocene cave deposits of Humpata, Southern Angola. Human Evolution, 7, 17–33. Wesselman, H. B., Black, M. T., & Asnake, M. (2008). Small mammals. In Y. Haile-Selassie & G. WoldeGabriel (Eds.), Ardipithecus kadabba: Late Miocene evidence from the Middle Awash, Ethiopia (pp. 105–134). Berkeley: University of California Press. White, J. A. (1991). North American Leporinae (Mammalia: Lagomorpha) from late Miocene (Clarendonian) to latest Pliocene (Blancan). Journal of Vertebrate Paleontology, 11, 67–89. Winkler, A. J. (2003). Rodents and lagomorphs from the Miocene and Pliocene of Lothagam, northern Kenya. In M. G. Leakey & J. M. Harris (Eds.), Lothagam, the dawn of humanity in eastern Africa (pp. 169–198). New York: Columbia University Press. Winkler, A. J., & Avery, M. (2010). Lagomorpha. In L. Werdelin & W. J. Sanders (Eds.), Cenozoic mammals of Africa (pp. 305–317). Berkeley: University of California Press.
Chapter 4
Macroscelidea Alisa J. Winkler
Abstract Two incomplete mandibles (plus a third tentatively referred) and an isolated P4 of Rhynchocyon pliocaenicus are reported from Localities 2, 3, and 10E, Upper Laetolil Beds, Laetoli, Tanzania, East Africa. The specimens are dated at ca. 3.7–3.6 Ma. Morphology of this newly recovered material confirms and enhances a diagnosis based previously on only the holotype and paratype. Mean jaw depth and almost all dental measurements of R. pliocaenicus are on average about 21% smaller than those of the extant species R. cirnei, R. petersi, and R. chrysopygus. Rhynchocyon pliocaenicus is diagnosed also by a p2 with a strong posterior heel with a prominent posterior basal cusp, consistent presence of an anterobuccal cingulum on p4-m2, and a posterior cingulum on p4-m1. The protoloph of P4 and M1 of R. pliocaenicus connects to the tip of the paracone. If the habitat preferences of R. pliocaenicus were similar to extant Rhynchocyon, then the presence of this fossil species suggests that closed canopy habitats were present at Laetoli ca 3.7–3.6 Ma, but probably relatively rare. Keywords Rhynchocyon • Pliocene • Laetoli • Sengis • Elephant shrews • Paleoecology
Introduction Paleontological collections from the Upper Laetolil and Upper Ndolanya Beds, Laetoli, Tanzania, by Mary Leakey and associates primarily from 1974 to 1982 yielded only two specimens of sengis (elephant shrews) (Butler 1987). This material was assigned to a new species of Rhynchocyon, R. pliocaenicus, which has been reported only from Laetoli. More extensive collecting in these units and in the Lower Laetolil, Upper Ndolanya and Ngaloba Beds by Terry A.J. Winkler (*) Roy M. Huffington Department of Earth Sciences, Southern Methodist University, Dallas, TX 75275, USA and Department of Cell Biology, University of Texas Southwestern Medical Center at Dallas, Dallas, TX 75390, USA e-mail:
[email protected] Harrison and associates (1998–2005) supports the rarity of sengis at Laetoli: only four new specimens were recovered out of a total sample of 15,019 specimens of all mammalian taxa from Laetoli (Harrison 2011). The new specimens of Rhynchocyon support Butler’s (1987) assignment of the Laetoli material to an extinct species, and necessitate only slight changes to the original diagnosis. Holroyd and Mussell (2005) have provided the most current classification and summary of the fossil record of the Macroscelidea. Sengis are known since the early Eocene. As fossils, and at present, the group is exclusively African. Paleogene reports (early or middle Eocene to early Oligocene) are only from North Africa and include representatives of two extinct subfamilies, the Herodotiinae (Hartenberger 1986; Simons et al. 1991; Tabuce et al. 2001) and Metoldobotinae (Schlosser 1910). The Miocene records are from Kenya, Uganda, and Namibia and include members of the Macroscelidinae (Stromer 1932; Butler 1984), Myohyracinae (extinct; Andrews 1914; Stromer 1922; Butler 1984; Senut 2003), and Rhynchocyoninae (Butler and Hopwood 1957; Butler 1969; Senut 2003). In the Plio-Pleistocene, the Macroscelidinae are reported from Tanzania (Butler and Greenwood 1976) and South Africa (Butler and Greenwood 1976), the Mylomygalinae (extinct) from South Africa (Broom 1948), and the Rhynchocyoninae from Tanzania (Butler 1987). At present, sengis are represented by the Macroscelidinae and the Rhynchocyoninae. Comparisons of the Laetoli sengi are made only within the Rhynchocyoninae, which include Rhynchocyon and the extinct genus Miorhynchocyon. Miorhynchocyon is known from the early Miocene of Kenya at Meswa Bridge, Koru, Legetet, Chamtwara, Songhor, Rusinga, Karungu, and Mfangano (Butler 1984). An isolated P3 and P4 are described from the middle Miocene at Fort Ternan, Kenya (Butler 1984). Miorhynchocyon is also present in the early Miocene at Napak, Uganda (personal observation) and in the middle Miocene at Arrisdrift, Namibia (Senut 2003). Four species of Miorhynchocyon have been described: M. clarki Butler and Hopwood, 1957, and M. rusingae Butler, 1969, are relatively well represented; M. meswae Butler, 1984, is known from a
T. Harrison (ed.), Paleontology and Geology of Laetoli: Human Evolution in Context. Volume 2: Fossil Hominins and the Associated Fauna, Vertebrate Paleobiology and Paleoanthropology, DOI 10.1007/978-90-481-9962-4_4, © Springer Science+Business Media B.V. 2011
67
68
single mandibular fragment with p3-p4, and M. gariepensis Senut, 2003, is represented by a mandible with p4-m2, a right m2, and an incomplete M1 (a skull and three isolated incisors are referred provisionally). The only fossil record of the genus Rhynchocyon is from Laetoli (Butler 1987). Schlitter (2005) lists three extant species of Rhynchocyon: R. chrysopygus from eastern Kenya; R. cirnei from Mozam bique, Malawi, Tanzania, Zambia, Democratic Republic of the Congo, and Uganda; and R. petersi from Tanzania and Kenya. A fourth species of Rhynchocyon, R. udzungwensis was described recently (Rovero et al. 2008) from the northern Udzungwa Mountains of Tanzania. At present, Rhynchocyon is “confined mainly to forest (lowland and montane) and thick riverine bush, although they have been taken in clearings amid grass and cane growth” (Nowak 1999: 1741). Kingdon (1974) notes that they occupy a wide range of elevation, from sea level to 2,300 m, and that they are dependent on shaded leaf litter.
A.J. Winkler
Rhynchocyon Peters, 1947 Rhynchocyon pliocaenicus Butler, 1987 (Figs. 4.1 and 4.2)
Methods Comparisons of the new material and casts of the holotype (rami and maxillary fragment with P4-M1) and paratype were made with dental remains of three extant species of Rhynchocyon, R. cirnei, R. petersi, and R. chrysopygus in collections of the Museum für Naturkunde der HumboldtUniversität, Berlin (ZMB), and the American Museum of Natural History, New York (AMNH) (Appendix). It is noteworthy that these three species were defined based on pelage and geographic distribution: cranial-dental characters were not used (Corbet and Hanks 1968; also see Kingdon 1974, who recognized only one species). Comparisons were not made with the four skulls of R. udzungwensis currently reported to be in museum collections (Rovero et al. 2008). Among fossil sengis, comparisons were made with all four species of Miorhynchocyon using published descriptions (descriptions only for M. gariepensis) and casts. A cast of the single specimen of M. meswae was unavailable; this specimen is a mandibular fragment with p3, p4, and the roots of p2. Casts are in collections of the Shuler Museum of Paleontology, Southern Methodist University, Dallas, Texas. Tooth terminology in this paper follows Butler (1987). Classification follows Holroyd and Mussell (2005) and Schlitter (2005).
Fig. 4.1 Rhynchocyon pliocaenicus from the Upper Laetolil Beds, Laetoli, Tanzania. (a) occlusal view of EP 552/01, left mandible with m1 and m2, tentatively referred. (b) EP 2743/00, right mandible with p4 (incomplete) and m1. (c–e) EP 2655/00, right mandible with p1, p3, m1, and m2; (c) occlusal view; (d) labial view; (e) lingual view. Anterior is to the left for all illustrations except (d). Lower scale applies to (c–e)
Systematic Paleontology ORDER Macroscelidea Butler, 1956 FAMILY Macroscelididae Bonaparte, 1838 SUBFAMILY Rhynchocyoninae Gill, 1872
Fig. 4.2 Rhynchocyon pliocaenicus from the Upper Laetolil Beds, Laetoli, Tanzania. Occlusal view of EP 637/03, isolated right P4. Anterior is to the right
4 Macroscelidea
Revised diagnosis: Mean jaw depth and dental measurements on average 21% less than Recent species of the genus. Second lower premolar with distinct posterior heel bearing a prominent posterior basal cusp; anterobuccal cingulum present on p4-m2; posterior cingulum present on p4-m1. Protoloph of P4 and M1 connects to tip of paracone versus continuing to the anterior cingulum. Holotype: LAET 75-2527, left mandible with base of p1, m1; right mandible with base p1, p2, p4-m2; five maxillae fragments, one with part of the left P4 (contra Butler 1987: fig. 4.1E) and also the M1. Considered one individual (Butler 1987). Paratype: LAET 79-5470, left mandible with p1-m1 (p4 missing talonid). Newly referred specimens: EP 2655/00, right mandible with p1, p3, m1, m2, and six associated bone fragments (two long bones, two vertebrae, one a caudal). EP 2743/00, right mandible with p4 (missing most of the paraconid), m1. EP 637/03, isolated right P4. EP 552/01, left mandible with m1 and m2 is tentatively referred. Stratigraphic horizon: The holotype and paratype were collected by M. Leakey and associates from the Upper Laetolil Beds: the holotype is from Loc. 2, and the paratype is from Loc. 5 (Butler 1987). Additional specimens were collected by T. Harrison and colleagues. These specimens are also from the Upper Laetolil Beds. EP 2655/00 and EP 637/03 are from Loc. 2, between Tuffs 5 and 7. EP 2743/00 is from Loc. 3, between Tuffs 7 and 8. EP 552/01 is from Loc. 10E, between Tuffs 5 and 7. All specimens were surface finds. Geologic age: ca. 3.7–3.6 Ma (Drake and Curtis 1987; Deino 2011). Description: The horizontal ramus of EP 2655, the most complete new specimen, is long and slender. The specimen is broken proximally where the ramus begins its gentle ascent. Mental foramina are present below p2 and p3, as noted by Butler (1987) for the hypodigm. A mental foramen was not observed below the talonid of m1, as on the holotype (Butler 1987). Morphology of the hypodigm is as described by Butler (1987) except as noted. In the original description of the species, Butler (1987) observed that p2 had a posterior heel. Comparison of the p2 (preserved only on the paratype and the right ramus of the holotype) to that of 16 extant specimens showed that R. pliocaenicus was distinctive in its strong development of the posterior heel on this tooth. Although some extant specimens were similar to the paratype in development, most were not. In particular, none of the extant material had a posterior basal cusp as tall as that on the right ramus of the holotype of R. pliocaenicus. Butler (1987) described his single p3 as having a protostylid. The single p3 in the new collections (EP 2655/00; Fig. 4.1c–e) lacks a protostylid: presence or absence of this cusp was variable in the
69
extant specimens examined. Butler (1987) also noted that the talonid of p3 was composed of two small cusps. This area appears to include a single cusp on EP 2655/00. Development of this area was variable in seven extant specimens. Butler (1987) did not have a complete P4 to describe. The newly recovered P4, EP 637/03 (Fig. 4.2), is roughly trapezoidal in outline. The anterior and posterior widths of the tooth are similar (3.44 and 3.60 mm), suggesting the tooth is a P4. This is in comparison to the greater discrepancy in anterior (3.15 mm) to posterior (2.35) width of the M1 (LAET 75-2527). As on the M1, the trigon is roughly “V” shaped with the base of the “V” pointing anteriorly. The protoloph connects laterally to the tip of the paracone. A paraconule is either lacking or might be considered minute and located at about the middle of the paraloph. There is a low cingulum anterior to the paracone. The protoconeparacone and metacone-hypocone pairs are essentially parallel and transverse, with the labial cusps slightly anterior. The major labial cusps are about two times the height of the lingual cusps. A curved crest connecting the metacone and hypocone closes the talon. There is a low posterolabial cingulum on the metacone. The trigon and talon basins are deep. The tooth has three roots: two labial and a single large lingual root. The teeth of EP 552/01 (only m1 and m2 are preserved and they are broken and heavily etched; Fig. 4.1a) are a maximum of about 28% (M1) to 8% (M2) smaller than the mean size for the species. Approximate measurements (in millimeters) of EP 552/01 (not included in the compiled measurements in Table 4.1) are: m1L = 2.42, m1Wtrigonid = 1.58, m1wtalonid = 1.50, m2L = 2.08, m2Wtrigonid = 1.58, and m2Wtalonid = 1.17. EP 552/01 also differs from others of the species in having a less distinct groove between the protoconid and paraconid, on m1 lacking a posterior cingulum behind the hypoconid, and on m2 having a narrower connection between the trigonid and talonid. However, the m1 and m2 both have distinct long and low anterobuccal cingula. Due to these differences, EP 552/01 is referred tentatively to R. pliocaenicus.
Discussion The new specimens of R. pliocaenicus and more extensive comparisons with extant species confirm Butler’s (1987) observations and provide additional information about R. pliocaenicus. Butler (1987) noted that R. pliocaenicus was smaller than Recent species of Rhynchocyon, but provided comparative measurements for only one specimen of R. cirnei (Butler 1987: Table 4.1). Comparative measurements for three extant species are provided here (Table 4.1), using a larger data set for R. cirnei (ten specimens), and an
70
A.J. Winkler Table 4.1 Measurement statistics for Rhynchocyon pliocaenicus and samples of extant R. cirnei, R. petersi, and R. chrysopygus. Measurements in mm Measurement
Taxon R. pliocaenicusa
R. cirneib, c
R. petersic
R. chrysopygusd
p1L N X SD OR
3 2.94 0.37 2.56–3.30
10 3.20 0.19 2.97–3.59
4 3.17 0.12 3.00–3.24
3 3.42 0.14 3.28–3.55
p1W N X SD OR
3 0.95 0.13 0.80–1.05
10 1.08 0.09 0.98–1.25
4 1.20 0.61 1.14–1.26
3 1.24 0.10 1.15–1.34
p2L N X SD OR
2 – – 2.60–2.80
10 3.24 0.24 2.88–3.70
4 3.26 0.25 2.96–3.56
3 3.22 0.10 3.11–3.30
p2W N X SD OR
2 – – 1.10–1.20
10 1.43 0.12 1.29–1.58
4 1.54 0.12 1.43–1.69
3 1.41 0.07 1.35–1.48
p3L N X SD OR
2 – – 2.84–3.35
10 4.08 0.27 3.74–4.60
4 4.37 0.38 4.00–4.86
3 4.33 0.42 4.08–4.82
p3W N X SD OR
2 – – 1.36–1.45
10 1.94 0.11 1.74–2.10
4 1.95 0.15 1.73–2.05
3 1.94 0.08 1.84–1.98
p4L N X SD OR
2 – – 3.95–4.08
10 4.98 0.57 4.35–6.00
4 5.25 0.24 5.03–5.59
3 5.10 0.20 4.88–5.26
p4Wtrigonid N X SD OR
3 2.04 0.04 2.00–2.08
10 2.60 0.20 2.40–2.98
4 2.66 0.11 2.60–2.83
3 2.69 0.12 2.56–2.79
p4Wtalonid N X SD OR
2 – – 2.20–2.24
10 2.79 0.19 2.54–3.10
4 2.95 0.17 2.80–3.14
3 2.97 0.02 2.94–2.98
m1L N X SD OR
4 3.44 0.08 3.35–3.55
10 3.97 0.20 3.69–4.27
4 4.06 0.14 3.91–4.24
3 4.27 0.14 4.10–4.36
m1Wtrigonid N X SD OR
4 2.13 0.14 1.92–2.24
10 2.72 0.14 2.43–2.85
4 2.74 0.08 2.66–2.83
3 2.86 0.09 2.75–2.92
m1Wtalonid N X SD OR
4 2.11 0.05 2.05–2.16
9 2.59 0.17 2.23–2.78
4 2.66 0.12 2.54–2.78
3 2.87 0.12 2.75–2.98
m2L N X SD OR
2 – – 2.24–2.25
10 2.72 0.27 2.40–3.20
4 2.71 0.21 2.57–3.01
2 – – 2.99–3.02 (continued)
4 Macroscelidea
71
Table 4.1 (continued) Measurement
Taxon R. pliocaenicusa
R. cirneib, c
R. petersic
R. chrysopygusd
m2Wtrigonid N X SD OR
2 – – 1.75–1.76
10 2.09 0.13 1.89–2.26
4 2.14 0.11 2.02–2.24
2 – – 2.17
m2Wtalonid N X SD OR
1 – – 1.25
9 1.58 0.11 1.41–1.79
4 1.65 0.10 1.58–1.79
2 – – 1.72–1.73
Jaw depth e N X SD OR
3 4.41 0.25 4.15–4.65
10 5.74 0.35 5.13–6.10
4 5.69 0.26 5.40–6.0
3 6.45 0.27 6.22–6.75
p1-m2L N X SD OR
2 – – 17.98–18.08
4 22.79 0.32 22.47–23.20
– – – –
2 – – 24.61–25.53
p4-m2L N X SD OR
1 – – 9.70
10 11.77 0.96 10.80–13.37
4 12.48 0.45 12.22–13.16
2 – – 12.85–13.14
P4L N X SD OR
1 – – 4.32
9 4.30 0.33 3.92–4.86
4 4.77 0.59 4.10–5.30
3 4.18 0.13 4.03–4.21
P4Wanterior N X SD OR
1 – – 3.44
9 4.02 0.33 3.70–4.63
4 4.43 0.31 4.15–4.87
3 4.41 0.04 4.37–4.44
P4Wposterior N X SD OR
2 – – 2.90–3.60
10 3.96 0.44 3.29–4.70
4 4.19 0.21 4.04–4.50
3 4.18 0.13 4.03–4.29
M1L N X SD OR
1 – – 3.15
10 3.82 0.24 3.42–4.20
4 3.88 0.44 3.55–4.53
3 4.02 0.07 3.95–4.08
M1Wanterior N X SD OR
1 – – 3.15
10 4.04 0.29 3.58–4.50
4 4.28 0.21 4.03–4.49
3 4.44 0.13 4.39–4.59
M1Wposterior N X SD OR
1 – – 2.35
10 3.28 0.23 2.96–3.52
4 3.36 0.18 3.26–3.63
3 3.56 0.14 3.40–3.67
L length, OR observed range, N number of specimens, SD standard deviation, W width, X mean a Combined measurements from Butler (1987) and the author’s measurements (maximum) on specimens from the Harrison collection. The latter were measured with an ocular on a Wild dissecting scope with measurement error of ± 0.08 mm b Measurements (maximum) of all extant specimens made with a Mitutoyo digital calipers with measurement error of ± 0.01 mm c Specimens of R. cirnei and R. petersi are from multiple localities d Specimens of R. chrysopygus are from the same locality e Lingual jaw depth at anterior end m1. Measured on the labial side when jaws were articulated
72
admittedly small data set for R. petersi (four specimens) and R. chrysopygus (three specimens). Rhynchocyon udzungwensis is larger than the other three extant species, for example, its mean weight is 25–50% greater than that of other species and its mean total body length is 10–20% longer (Rovero et al. 2008). In almost all dental measurements, R. pliocaenicus is smaller than Recent species: mean measurements for R. pliocaenicus are on average 21% smaller (range 3% larger to 34% smaller) than mean measurements for the extant species. Only the length of P4 (one specimen of R. pliocaenicus) is comparable (R. cirnei) or slightly larger (by 3% to R. chrysopygus) than the mean length of extant species. P4 length of R. pliocaenicus is 9% shorter than the mean length for R. petersi. The P4 of R. pliocaenicus is narrower than the mean values for extant species, but the posterior width of the larger Laetoli P4 is within the range of variation for R. cirnei. Although the mean length and width of the p1 of R. pliocaenicus are less than the mean values for the three extant species, the length of the p1 of R. pliocaenicus is within the range of variation seen in the other species, and the width is within the range of variation for R. cirnei (but not the other extant species). As noted by Butler (1987), an anterobuccal cingulum is present on the paraconid of p4-m2 of R. pliocaenicus. Butler observed this on only a few specimens of Recent species. In the present study, this structure was seen distinctly on the m1 and only faintly on the m2 (and not at all on p4) on only 2 out of 16 extant specimens of Rhynchocyon. A posterior cingulum is present consistently on the p4-m1 of R. pliocaenicus. Butler did not observe this structure on extant material. In the present study, a posterior cingulum was observed on 1 of 16 recent specimens. Butler (1987) noted that on the one known M1 of R. pliocaenicus (LAET 75-2527) the protoloph was a strong crest connecting to the tip of the paracone. In the present study, Butler’s (1987) observation that the protoloph of M1 of Recent Rhynchocyon usually does not connect to the tip of the paracone was confirmed. In nine of ten Recent specimens, the protoloph of M1 extends anterior to the paracone. On one of those specimens (R. petersi, ZMB 20025, on one maxilla, but not the other), the protoloph did turn toward and contact the paracone, but the bend was not as abrupt as seen on the fossil. On the P4 of R. pliocaenicus (EP 637/03), the protoloph connects to the tip of the paracone. This condition is variable on the five Recent specimens examined: On two specimens the protoloph extended anterior to the paracone and on three specimens the protoloph contacted the paracone on one side of the skull but not on the other. Butler (1987) also made comparisons of R. pliocaenicus with the early and middle Miocene genus Miorhynchocyon (M. clarki and M. rusingae). Comparisons are extended here to M. meswae and M. gariepensis. Butler’s (1987: 86)
A.J. Winkler
observations are augmented here by comparisons with the larger sample of R. pliocaenicus. Miorhynchocyon was diagnosed (for the dentition) as differing from Rhynchocyon in the following (Butler 1984): (1) the oblique crest on the lower molariform teeth (anterior hypoconid crest) ends midway between the protoconid and hypoconid (in extant Rhynchocyon it joins the metaconid), (2) a metastylid is absent on dp4 and m1, (3) on p4 and m1 the paraconid is higher and located more lingually, and (4) the cheek teeth are more brachydont. Butler noted that in size, measurements of R. pliocaenicus fell within the range of values for M. clarki (Butler 1984: table 2). The additional specimens of R. pliocaenicus demonstrate that it is comparable in size or slightly larger than M. clarki. Rhynchocyon pliocaenicus is smaller than M. rusingae, comparable to slightly smaller than the single specimen of M. meswae, and comparable (p4, m1) to smaller (m2) than M. gariepensis (Senut 2003). Butler (1987) observed that on the p4-m2 of R. pliocaenicus the anterior hypoconid crest connected the hypoconid to the metaconid: it did not end midway between the hypoconid and metaconid as in Miorhynchocyon. Connection of this crest to the metaconid was observed on the additional specimens of R. pliocaenicus, although it was less clear on the m2 of EP 2655/00. The dp4 is not known for R. pliocaenicus, but the m1 lacks a metastylid, comparable to Miorhynchocyon. The paraconids of p4 and m1 are usually low, but the teeth of R. pliocaenicus are generally less brachydont compared to Miorhynchocyon. As noted by Butler (1987), R. pliocaenicus resembles Miorhynchocyon in development of the anterobuccal and posterior cingula (also described as present on m1 of M. gariepensis, but not described or illustrated for p4 [Senut 2003]), and in the M1 having the protoloph connecting to the tip of the paracone (although Miorhynchocyon usually lacks a paraconule). The incomplete M1 of M. gariepensis is not illustrated and the most anterior portion of the tooth is not discussed. The new P4 of R. pliocaenicus is also similar to the P4 of Miorhynchocyon in having the protoloph connecting to the tip of the paracone (on R. pliocaenicus the paraconule is either lacking or minute; Miorhynchocyon usually lacks a paraconule). Butler (1987) also commented on differences in the presence or absence of a protostylid on p2 and p3. In all species of Rhynchocyon and M. clarki, p2 lacks a protostylid. A protostylid is present on the p2 of M. rusingae (condition of M. meswae unknown). A protostylid is absent on the p3 of M. clarki, present on M. rusingae and M. meswae (Butler 1984), present on R. pliocaenicus, and generally present in recent species (Butler 1987). Presence or absence of protostylids on p2 and p3 thus does not appear to be a useful character for differentiating between the two genera. In summary, R. pliocaenicus differs from Miorhynchocyon primarily in characters that differentiate the two genera,
4 Macroscelidea
but R. pliocaenicus still shares some characters with Miorhynchocyon that differentiate both of these taxa from extant species of Rhynchocyon.
Conclusions Sengis remain an extremely rare component of the Laetoli fauna in spite of eight seasons (and over 15,000 specimens) of additional collecting. From both the Leakey and Harrison collections, Rhynchocyon pliocaenicus is known from only the Upper Laetolil Beds, and not the Upper Ndolanya or Lower Laetolil Beds. It should be noted that for the Harrison collections, all specimens were collected as surface finds and collectors recovered all specimens they considered anatomically identifiable (Su and Harrison 2008). Specimens from the Leakey collections, Upper Laetolil Beds, were collected primarily as surface finds (except most rodents were recovered from screen washing at Localities 5 and 6), but excavations yielded many of the fossils from the Upper Ndolanya Beds (Leakey 1987). The Upper Ndolanya Beds are well sampled and have yielded an extensive fauna. However, there is a strong taphonomic bias against small mammals in this unit, which may be reflected in the lack of sengis (Harrison, personal communication). The Lower Laetolil Beds have not yet been sampled adequately (i.e., the collection from these beds is about 50× smaller than that from the upper unit) and this sampling bias may account for the lack of sengis (Harrison, personal communication). Small mammals do comprise a large proportion of the fauna from the Upper Laetolil Beds. For example, in the Harrison collection 37.5% of all specimens of mammals are lagomorphs (Harrison, personal communication): lagomorphs represent about 31% of the Leakey collection (Leakey 1987: Fig. 1.3). Although not as common as lagomorphs, rodents are well represented in both the Leakey (about 6%; Leakey 1987: Fig. 1.3; Denys 1987) and Harrison collections (6.9%; Harrison, personal communication). Although smaller in size than the Laetoli lagomorph, Serengetilagus, Rhynchocyon pliocaenicus is larger than many of the rodent taxa reported from the Upper Laetolil Beds (however, 70% of all rodents were spring hares, Pedetes, which are larger than Rhynchocyon; Su and Harrison 2008). Thus, it is unlikely that the rarity of sengis in the Upper Laetolil Beds is from either a preservational or collecting bias against small mammals in general. However, relatively little screen washing was done by both the Leakey and Harrison teams, and that may account for the small proportion of the smallest small mammals. The paucity of sengis from the Upper Laetolil Beds likely reflects the scarcity of suitable paleohabitats. It is, admittedly, a tenuous assumption assuming that the habitat preferences of
73
a fossil species were similar to those of its extant relatives. However, the dental remains (the only part of the animal preserved) of R. pliocaenicus are very similar to those of extant species, so it is likely that their diets, and perhaps their preferred habitat, were similar. Leakey (1987) suggested that the Laetolil Beds represented a dry savanna habitat. Andrews’ (2006: 572) interpretation was that of a “heavily wooded environment with local patches of forest and few open grassland areas.” Su and Harrison’s (2008) interpretation was more similar to that of Andrews’, but they considered the Upper Laetolil Beds to be derived from a predominately open woodland, which also included extensive open bushland, shrubland, and grassland habitats. Recent species of Rhynchocyon prefer areas of closed canopy and are dependent on shaded leaf-litter. This suggests that closed canopy habitats were present, but rare, at Laetoli ca. 3.7–3.6 Ma. It is noteworthy that the only published fossil record of Rhynchocyon is from Laetoli, and the only other published records of the Rhynchocyoninae are 84 specimens from nine early Miocene faunas (plus two isolated teeth from the middle Miocene; Butler 1984) from Kenya and three (possibly seven) specimens from the middle Miocene of Namibia (Senut 2003). East African early Miocene sites are considered to sample a larger percentage of closed canopy habitat than Laetoli, and this is likely reflected in the relatively greater abundance of Rhynchocyoninae in the earlier samples. Acknowledgments I am grateful to T. Harrison for the invitation to study the Laetoli sengis and for discussions about the overall fauna and paleoenvironments at Laetoli. Robert Asher (formerly at the ZMB) and E. Westwig (AMNH) kindly provided access to extant comparative material. I thank D. Winkler for photographing the specimens. Dale Winkler, Pat Holroyd, and two anonymous referees provided constructive reviews of the manuscript. Funding for fieldwork at Laetoli was provided by grants to T. Harrison from the National Geographic Society, the Leakey Foundation, and the National Science Foundation (grants BCS-9903434 and BCS-0309513). The latter grant also provided travel funds to Berlin and New York City for A. Winkler.
Appendix 4.1 Species and provenance of comparative specimens of extant Rhynchocyon Specimen Taxon number Provenance R. cirnei
R. cirnei R. cirnei R. cirnei macrurus R. cirnei hendersoni
R. petersi
ZMB 19987, 19992, 19995 ZMB 20014 ZMB 31798 BMNH 63.1854 AMNH 81332– 81335 ZMB 11428
Nynga, Bez Songea, Tanzania Mitononi am Mbemkuru, Tanzania Mikindani, Tanzania Not given (Butler 1987: table 4.1) Rungwe, Tanzania
Majoni, Tanzania (continued)
74
A.J. Winkler
Appendix 4.1 (continued) Specimen Taxon number R. petersi R. petersi R. chrysopygus
ZMB 20023, 20025 ZMB 84895 AMNH 187231, 187232, 187234
Provenance Mitononi am Mbemkuru, Tanzania Kimbuguru, Pangani Bez, Tanzania Kilifi District, Kenya
References Andrews, C. W. (1914). On the lower Miocene vertebrates from British East Africa collected by Dr. Felix Oswald. Quarterly Journal of the Geological Society of London, 70, 163–186. Andrews, P. (2006). Taphonomic effects of faunal impoverishment and faunal mixing. Palaeogeography, Palaeoclimatology, Palaeoecology, 241, 572–589. Broom, R. (1948). Some South African Pliocene and Pleistocene mammals. Annals of the Transvaal Museum, 21, 1–38. Butler, P. M. (1969). Insectivores and bats from the Miocene of East Africa: New material. In L. S. B. Leakey (Ed.), Fossil vertebrates of Africa (Vol. 1, pp. 1–37). London: Academic. Butler, P. M. (1984). Macroscelidea, Insectivora and Chiroptera from the Miocene of East Africa. Palaeovertebrata, 14(3), 117–200. Butler, P. M. (1987). Fossil insectivores from Laetoli. In M. D. Leakey & J. M. Harris (Eds.), Laetoli: A Pliocene site in northern Tanzania (pp. 85–87). Oxford: Clarendon. Butler, P. M., & Greenwood, M. (1976). Lower Pleistocene elephantshrews (Macroscelididae) from Olduvai and Makapansgat. In R. J. G. Savage & S. C. Coryndon (Eds.), Fossil vertebrates of Africa (Vol. 4, pp. 1–56). London: Academic. Butler, P. M., & Hopwood, A. T. (1957). Insectivora and Chiroptera from the Miocene rocks of Kenya colony. Fossil Mammals of Africa, 13, 1–35. Corbet, G. B., & Hanks, J. (1968). A revision of the elephant-shrews, family Macroscelididae. Bulletin of the British Museum (Natural History), Zoology, 16, 47–111. Deino, A. L. (2011). 40Ar/39Ar dating of Laetoli, Tanzania. In T. Harrison (Ed.), Paleontology and geology of Laetoli: Human evolution in context (Geology, geochronology, paleoecology and paleoenvironment, Vol. 1, pp. 77–97). Dordrecht: Springer. Denys, C. (1987). Fossil rodents (other than Pedetidae) from Laetoli. In M. D. Leakey & J. M. Harris (Eds.), Laetoli: A Pliocene site in northern Tanzania (pp. 118–170). Oxford: Clarendon. Drake, R., & Curtis, G. (1987). K–Ar geochronology of the Laetoli fossil localities. In M. D. Leakey & J. M. Harris (Eds.), Laetoli: A Pliocene site in northern Tanzania (pp. 48–52). Oxford: Clarendon.
Harrison, T. (2011). Laetoli revisited: Renewed palaeontological and geological investigations at localities on the Eyasi Plateau in northern Tanzania. In T. Harrison (Ed.), Paleontology and geology of Laetoli: Human evolution in context (Geology, geochronology, paleoecology and paleoenvironment, Vol. 1, pp. 1–15). Dordrecht: Springer. Hartenberger, J.-L. (1986). Hypothèses paléontologiques sur l’origine des Macroscelidea (Mammalia). Comptes rendus de l’Académie des Sciences, Sciences de la Terre et des Planètes B302, Série II, Paris, 5, 247–249. Holroyd, P. A., & Mussell, J. C. (2005). Macroscelidea and Tubulidentata. In K. D. Rose & J. D. Archibald (Eds.), The rise of placental mammals (pp. 71–83). Baltimore: John Hopkins University Press. Kingdon, J. (1974). East African mammals, an atlas of evolution in Africa. Vol. IIA, Insectivores and bats. Chicago: University of Chicago Press. Leakey, M. D. (1987). Introduction. In M. D. Leakey & J. M. Harris (Eds.), Laetoli: A Pliocene site in northern Tanzania (pp. 1–22). Oxford: Clarendon. Nowak, R. M. (1999). Mammal species of the world (3rd ed., Vol. 1). Baltimore: John Hopkins University Press. Rovero, F., Rathbun, G. B., Perkins, A., Jones, T., Ribble, D. O., Leonard, C., Mwakisoma, R. R., & Doggart, N. (2008). A new species of giant sengi or elephant-shrew (genus Rhynchocyon) highlights the exceptional biodiversity of the Udzungwa Mountains of Tanzania. Journal of Zoology, 274, 126–133. Schlitter, D. A. (2005). Order Macroscelidea. In D. E. Wilson & D. M. Reeder (Eds.), Mammal species of the world. A taxonomic and geographic reference (3rd ed., Vol. 1, pp. 82–85). Baltimore: John Hopkins University Press. Schlosser, M. (1910). Über einige fossile Säugetiere aus dem Oligocan von Ägypten. Zoologischer Anzeiger, 35, 500–508. Senut, B. (2003). The Macroscelididae from the Miocene of the Orange River, Namibia. In M. Pickford & B. Senut (Eds.), Geology and palaeobiology of the central and southern Namib (Palaeontology of the Orange River Valley, Vol. 2, pp. 119–141). Windhoek: Geological Survey of Namibia, Memoir 19. Simons, E. L., Holroyd, P. A., & Brown, T. M. (1991). Early Tertiary elephant shrews from Egypt and the origin of the Macroscelidea. Proceedings of the National Academy of Sciences of the United States of America, 88, 9734–9737. Stromer, E. (1922). Erste Mitteilung über tertiäre Wirbeltier-Reste aus Deutsch-Südwestafrika. Sitzungsberichte Mathematik und Physik Klasse Bayern Akademische Wissenschaft, 1921, 331–340. Stromer, E. (1932). Paleothentoides africanus nov. gen., nov. spec., ein erstes Beuteltier aus Afrika. Sitzungsberichte der MathematischNaturwissenschaftlichen Klasse der Bayerischen Akademie der Wissenschaften, 1931, 177–190. Su, D. F., & Harrison, T. (2008). Ecological implications of the relative rarity of fossil hominins at Laetoli. Journal of Human Evolution, 55, 672–681. Tabuce, R., Coiffait, B., Coiffait, P.-E., Mahboubi, M., & Jaeger, J.-J. (2001). A new genus of Macroscelidea (Mammalia) from the Eocene of Algeria: A possible origin for elephant-shrews. Journal of Vertebrate Paleontology, 21, 535–546.
Chapter 5
Galagidae (Lorisoidea, Primates) Terry Harrison
Abstract An additional specimen of a fossil galagid was recently recovered from the Upper Laetolil Beds at Laetoli in northern Tanzania. This new find represents the most complete specimen of a galagid known from Laetoli, and comprises associated partial right and left mandibular corpora. The galagid material from Laetoli can all be attributed to a single species, previously referred to as Galago sadimanensis. However, the taxon is sufficiently distinct from all extant galagids, as well as stem galagids from the Miocene of East Africa, to be placed in its own genus, Laetolia. The fossil record of galagids from the Pliocene of Africa is exceedingly poor, and Laetolia sadimanensis represents the best-known form. Laetolia can be distinguished from other galagids by its unique suite of morphological features. The stout and vertical implantation of P2, the steeply inclined and robust symphysis, and the relatively deep corpus are all specialized features that are probably functionally linked. However, Laetolia has a less molariform P4 than extant galagids, and it can be inferred to represents their primitive sister taxon. Based on molecular clock estimates, extant galagids shared a last common ancestor during the late Oligocene. It is interesting, therefore, to discover a sister taxon of extant galagids surviving in East Africa until at least the Pliocene, contemporary with more advanced crown members of the clade. From a paleoecological perspective, the occurrence of fossil galagids at Laetoli implies the presence of habitats with at least a sparse coverage of trees and/or thorn bush. Keywords Galagids • Laetoli • Mabaget Formation • Pliocene • Phylogeny
Introduction A single species of galagid, Galago sadimanensis, is represented by a number of partial mandibles from the Upper Laetolil Beds (3.63–3.85 Ma) at Laetoli (Walker 1987; T. Harrison (*) Center for the Study of Human Origins, Department of Anthropology, New York University, 25 Waverly Place, New York, NY 10003, USA e-mail:
[email protected] Harrison 2010). A mandibular fragment from the Mabaget Formation (~5 Ma), in the Tugen Hills of Kenya, collected earlier, has been referred to the same species (Walker 1987). Renewed investigations at Laetoli have yielded an additional galagid specimen (EP 1064/03). The specimen was discovered by Chris Robinson in 2003 at Loc. 10W in the Upper Laetolil Beds between Tuffs 1 and 3 (~3.8 Ma). It consists of associated right and left mandibular fragments with P2-M3 and P2, P4-M1 respectively, and it represents the most complete specimen of Galago sadimanensis known. The aim of this chapter is to describe briefly the new specimen, to present an updated account of the morphology of G. sadimanensis to highlight its distinctive features, and to clarify its taxonomic and phylogenetic relationships. As discussed below, the species is considered to be sufficiently distinct from extant galagids and from other fossil genera to be placed in its own genus. The fossil record of galagids from the Plio-Pleistocene is exceedingly poor (Harrison 2010). Apart from Galago sadimanensis, the only other extinct species formally described is Otolemur howelli, based on a fragmentary maxilla, an isolated M2, and an edentulous mandible from the lower part of the Shungura Formation (~3.0–3.2 Ma) in the Omo Valley, Ethiopia (Wesselman 1984). Fragmentary finds of other fossil galagids are known from localities in East Africa, some of which probably belong to extant taxa. Several mandibular fragments, isolated teeth, and postcranial elements from Bed I (~1.8 Ma), Olduvai Gorge, northern Tanzania, can be referred to the extant species, Galago senegalensis (Simpson 1965; Szalay and Delson 1979; Gebo 1986; Harrison 2010). Wesselman (1984) described a fragmentary M2 from lower Member G (~2.0 Ma) of the Shungura Formation, Omo, Ethiopia, which he referred to Galago senegalensis, but the tooth is smaller than those of the modern taxon and it is best considered an indeterminate species (Harrison 2010). Wesselman (1984) also described an isolated M2 from upper Member B of the Shungura Formation (~3.0 Ma), which is very similar to Galagoides zanzibaricus, except that the crown is slightly narrower. Denys (1987) reported an isolated upper canine of a galagid from the Humbu Formation (~1.3–1.7 Ma) at Peninj, Tanzania, which is consistent in
T. Harrison (ed.), Paleontology and Geology of Laetoli: Human Evolution in Context. Volume 2: Fossil Hominins and the Associated Fauna, Vertebrate Paleobiology and Paleoanthropology, DOI 10.1007/978-90-481-9962-4_5, © Springer Science+Business Media B.V. 2011
75
76
morphology and only slightly smaller than that of extant Galago senegalensis. Finally, Harris et al. (2003) described a mandibular fragment with M2 of a diminutive galagid from Kanapoi in Kenya (~4.1–4.2 Ma). Fossil galagids are also known from the late Miocene of Africa. These include an isolated upper molar from Harasib 3a (~9–10 Ma) in Namibia (Conroy et al. 1993, 1996; Rasmussen and Nekaris 1998), several isolated teeth and postcranials of a small galagid, Galago farafraensis, from Sheikh Abdallah (~10–11 Ma) in Egypt (Pickford et al. 2006), and an undescribed mandible of a galagid from Kapsomin, in the Lukeino Formation (~6 Ma) in Kenya (Mein and Pickford 2006). Extant galagos are included together in a single family, the Galagidae, which is restricted to sub-Saharan Africa. There are at least 24 species currently recognized, belonging to five genera - Galago, Galagoides, Otolemur, Euoticus, and Sciurocheirus (Kingdon 1997; Bearder 1999; Masters and Bragg 2000; Groves 2001; Grubb et al. 2003). A further genus name might be required to accommodate the Galagoides orinus group (sensu Grubb et al. 2003) if the Galagoides demidovii group proves not to be its sister taxon (see Fabre et al. 2009). The extant members of the family are characterized by the following cranio-dental features: molariform upper and lower P4; upper molars with large hypocone on an expanded distolingual lobe, well-developed prehypocone crista, deeply notched distal margin, long and distobuccally directed postmetacrista; lower molars with, elongated subtriangular trigonid with beak-like mesial margin; relatively lightly constructed cranium; orbits lacking strong frontation and raised margins; shallow mandible and lower face; very inflated auditory bulla with pneumatization extending into the mastoid region (Harrison 2010).
T. Harrison
P2 stouter, lower-crowned, and more vertically implanted; P2 larger in occlusal area than P4; P3 and P4 relatively small in relation to M1; P4 less molarized with shorter and more ovoid crown, rounded mesial margin, less well-developed metaconid, shorter and narrower talonid basin, and weakly developed entoconid and hypoconid; lower molars relatively narrower (Emended from Walker 1987; Harrison 2010). It differs from Komba (early and middle Miocene of East Africa) in the following respects: relatively thicker mandibular corpus; mandibular symphysis more vertical, with a relatively greater cross-sectional area, and an inverted tear-drop sagittal section; P2 relatively larger; P4 relatively shorter, with a less pronounced mesial beak, a less well-developed metaconid, relatively smaller distal cuspules, a shorter, broader and shallower talonid basin; P4 slightly larger in occlusal area relative to M1. Differs from Progalago (early Miocene of East Africa) in having a relatively shallower mandibular corpus that does not increase in depth posteriorly, and lacks a flange-like inferior margin; mandibular symphysis steeper and more robust; P2 relatively larger; P4 relatively shorter, with more pronounced distal cuspules, and a smaller talonid basin; P4 slightly larger in occlusal area relative to M1; lower molars narrower, with less-pronounced buccal flare, longer and more triangular trigonid, more pronounced mesial beak, greater height differential between the trigonid and talonid, narrower and shallower talonid basin, more voluminous cusps, weaker occlusal crests, and more obliquely oriented distal margin (Walker 1987; Phillips and Walker 2002; Harrison 2010). Type species: Laetolia sadimanensis (Walker 1987). Included species: L. sadimanensis (Walker 1987). Holotype: LAET 74-294, right mandibular fragment with P2-M2. Laetoli, Tanzania. Hypodigm: Specimens listed in Table 5.1, plus KNM-BC 1646 from the Mabaget Formation, Kenya.
Systematics Order Primates Linnaeus, 1758 Suborder Strepsirrhini Geoffroy, 1812 Infraorder Lorisiformes Gregory, 1915 Superfamily Lorisoidea Gray, 1821 Family Galagidae Gray, 1825 Subfamily Galaginae Gray, 1825 Genus Laetolia gen. nov. Diagnosis: A galagid similar in overall dental dimensions to the extant Galago senegalensis. It differs from extant genera of galagids (i.e., Galago, Euoticus, Galagoides, Sciurocheirus and Otolemur) in the following features: relatively deeper and more robust mandibular corpus; mandibular symphysis more vertical, with a relatively greater cross-sectional area, and an inverted tear-drop (rather than oval) sagittal section;
Table 5.1 List of galagid specimens from the Upper Laetolil Beds, Laetoli Specimen
Locality
Element
LAET 74-294
Loc. 5
LAET 75-2433
Loc. 10W
LAET 75-2880
Loc. 10W
LAET 76-4144
Loc. 11
LAET 78-4702
Loc. 7
EP 1064/03
Loc. 10W
Left mandibular fragment with P2-M2 Right mandibular corpus and much of the ramus with P2-M2 Left mandibular fragment with P2-P3 Left mandibular fragment with base of P2 Right mandibular fragment with M2-M3 Right mandibular corpus with P2-M3 and left mandibular corpus with P2, P4-M1
5 Laetoli Galagids
Distribution: Pliocene, ~3.6–5.0 Ma. Upper Laetolil Beds, Laetoli, Tanzania and Mabaget Formation, Kapchebrit, Baringo Basin, Kenya. Diagnosis: Same as genus.
Description of EP 1064/03 The specimen consists of two associated mandibular fragments comprising the right mandibular corpus with P2-M3 and the left mandibular corpus with P2 and P4-M1 (Fig. 5.1). The right mandibular fragment comprises the entire corpus and the anterior and inferior aspects of the ramus. The corpus is preserved anteriorly as far as the symphysis, but the alveoli for the canine and incisors are poorly preserved, and the symphysis is incomplete superiorly. The rest of the corpus is entire and well-preserved, except for some faint pitting caused by weathering. On the lateral side of the corpus below P4 there is single large elliptical mental foramen. A tiny accessory foramen is located vertically below P3. The ramus is broken obliquely and abraded, preserving only the root of the anterior margin of the ramus and the inferior border, extending posteriorly 8.3 mm beyond M3. The dentition is lightly worn and generally well-preserved, except for some minor weathering and abrasion. P2 is missing the tip of the crown.
Fig. 5.1 EP 1064/03, right and left mandibular fragments of Laetolia sadimanensis. (a) right mandibular fragment with P2-M3, lateral view; (b) right mandibular fragment with P2-M3, medial view; (c) right mandibular fragment with P2-M3, occlusal view; (d) left mandibular
77
The left mandibular fragment consists of the symphyseal region and the corpus as far posteriorly as the alveolus for the anterior root of M2. The symphysis and alveoli for the canine and incisors are better preserved than on the right side and are almost complete. A large mental foramen is located below P3/P4, and a minute accessory foramen is positioned below P2. There is a fresh break through the corpus behind M1, indicating that the posterior portion of the corpus was detached after the specimen eroded out onto the surface. The preserved teeth are complete, but their enamel surfaces are slightly weathered. P3 is represented by the roots only.
Morphology of Laetolia sadimanensis Only the lower dentition and mandibles of Laetolia sadimanensis are known (see Table 5.1). The mandibular corpus is relatively deep and more robust than in modern galagids. It maintains a constant depth below the cheek teeth or shallows slightly posteriorly. There is a single mental foramen positioned vertically below P3 or P4 (Table 5.2), and located just below mid-height (40–45% up from the inferior margin). A tiny accessory foramen is commonly located just anterior to the main foramen. The symphysis is stout, with an anteroposterior thickness of 65–75% of its height, compared with
fragment with P2, P4-M1, lateral view; (e) left mandibular fragment with P2, P4-M1, medial view; (f) left mandibular fragment with P2, P4-M1, occlusal view. All to the same scale
78
T. Harrison
55–65% in extant galagids. The symphysis was unfused with no indication of synostosis. In section, the symphysis forms the shape of an apostrophe, with an almost circular superior portion and a smaller inferior torus that projects posteriorly. In modern galagids, the symphysis forms a flat ellipse, with no development of an inferior torus. The symphysis is more steeply inclined than in modern galagids, as well as in Miocene taxa, with a mean angle of the long-axis at 66° to the alveolar plane. The incisors and canines are not preserved in any of the specimens, so it is not possible to determine the degree of procumbency of the toothcomb. The average minimum width between the P2s can be estimated to be 3.2 mm (range 3.0–3.6 mm). This is comparable in breadth to the similar-sized Galago senegalensis, and implies that the Laetoli galagid had a narrow toothcomb as in extant taxa. Parts of the ramus are preserved in LAET 75-2433, LAET 78-4702 and EP 1064/03. The anterior margin of the ramus is set far back from M3 and inclined posteriorly at angle of about 125° relative to the alveolar plane as in extant galagids. The coronoid process is not preserved. The base of the condyle is preserved in LAET 75-2433, and appears to have been slightly lower than in extant galagids, being situated just above the level of the occlusal plane of the molars. The ramus is antero-posteriorly quite long, being 185% the length of the molar row. This exceeds the relative length in extant
galagids, and more closely approximates the condition seen in some lorisids, such as Perodicticus. The posterior angle of the mandible is not preserved, but judging from the strongly downturned inferior margin behind M3 it was quite expanded. A similar pattern is seen in Otolemur, but is less pronounced in the smaller extant galagids. Dimensions of the lower cheek teeth of Laetolia sadimanensis are presented in Table 5.3. P2 is a robust caniniform tooth, relatively vertically implanted, with a single stout root. It has a convex mesial face and a longer concave distal face, with a short distal heel. The lingual face is bordered basally by a narrow cingulum. The P2 is more robust than the similar-sized tooth in Galago senegalensis, and contrasts with the more procumbent sectorial tooth seen in all extant galagids. Even accounting for the variability in the form of P2 in modern galagids, in which the larger species tend to have the most vertical and caniniform teeth, the degree of procumbency in the Laetoli galagid is much less marked even than in Otolemur. In the robusticity and orientation of the P2 Laetolia sadimanensis approaches the specialized condition in extant lorisids. P2 is larger than P4, with the average occlusal area 117% of that of P4. In Miocene and extant galagids, the occlusal area of P2 is typically smaller than P4 (e.g., Progalago dorae, 97%; Komba robustus, 88%; Komba winamensis, 67%; Galago senegalensis,
Table 5.2 Mandibular dimensions (mm) of Laetolia sadimanensis LAET 294
LAET 2433
LAET 2880
LAET 4144
LAET 4702
KNM-BC 1646
EP 1064/03 (right)
EP 1064/03 (left)
Angle of symphysisa 64° 64° 71° 63° 70° 62° Depth at symphysisb 5.2 5.7 5.5 5.2 Thickness at symphysisc 3.4 4.2 3.6 3.6 3.3 Depth at M1d 5.1 5.2 5.1 4.0 4.9 4.7 Depth at M2d 4.9 4.9 5.0 4.4 Depth at M3d 5.4 5.2 4.5 Position of foramene mid P4 mes P4 dist P3 mid P4 mes P4 mes P4 P3/P4 a Angle of the symphysis midline axis relative to the alveolar plane of the mandibular corpus b Maximum length of the symphyseal face measured along the midline of its long-axis c Maximum breadth of the symphyseal face measured perpendicular to the midline long-axis d Infero-superior depth of the mandibular corpus below the lower molars e Vertical position of the main mental foramen below the cheek teeth: dist P3, below the distal moiety of P3; P3/P4, below the contact between P3 and P4; mes P4, below the mesial moiety of P4; mid P4, below the transverse midline of P4 Table 5.3 Dental dimensions (mm) of Laetolia sadimanensis Specimen LAET 74-294 LAET 75-2433 LAET 75-2880 LAET 76-4144 LAET 78-4702 KNM-BC 1646 EP 1064/03 (right) EP 1064/03 (left) Mean
P2 MD
BL
2.1 2.4 2.2 2.2 2.4 2.0 2.3 2.2
P3 MD
BL
1.4 1.6 1.4 1.3
1.8 2.0 1.9
0.9 1.0 1.0
1.3 1.3 1.4 1.4
1.8 1.8
1.1 1.0
1.9
1.0
P4 MD 2.0 2.0
1.7 1.7 1.9
BL
M1 MD
BL
1.3 1.5
2.0 2.0
1.8 1.9
1.8 2.2 2.1 2.0
1.5 1.8 1.8 1.8
1.4 1.4
M2 MD
BL
M3 MD
BL
2.0
2.3
2.0
2.8
1.8
2.1
1.7
2.2
1.4
2.2
1.8
2.5
1.6
5 Laetoli Galagids
68%; Otolemur crassicaudatus, 96%; Galagoides zanzibaricus, 68%). In this respect, Laetolia sadimanensis begins to approach the more specialized condition in extant lorisids (e.g., Loris tardigradus, 137%; Perodicticus potto, 136%; Nycticebus coucang, 181%). P3 is a long, slender sectorial tooth with a single main cusp, the protoconid, situated in the midline one-third back from the mesial margin of the crown. The mesial and distal crests are sharp. The distolingual crest is more rounded. There is a weakly developed lingual cingulum. There are two roots. The tooth is similar in overall morphology to that of G. senegalensis, but differs in being relatively smaller in relation to M1. The average occlusal area of P3 is 53% of that of M1 (compared with 63% in Galago senegalensis). P4 is a short, ovoid tooth with an elevated protoconid and a poorly developed metaconid. The talonid basin is short, and bordered distally by a pair of low, rounded tubercles, the entoconid and hypoconid. There are two roots. P4 is relatively small in relation to M1. The occlusal area averages 70% of that of M1, slightly greater than in early Miocene galagids (e.g., Progalago dorae, 64%; Komba robustus, 63%), but smaller than in extant taxa (e.g., Otolemur crassicaudatus, 74%; Galago senegalensis, 77%). In sum, the P4 is less molarized than in extant galagids, with a shorter, more ovoid crown, less well-developed metaconid, shorter and narrower talonid basin, and weakly developed entoconid and hypoconid. Laetolia also differs from extant galagids in having a less pronounced prow-like mesial beak at the front of the tooth. The P4 of Laetolia is more derived than Komba from the Miocene of East Africa, in having a broader talonid basin and in being relatively larger in comparison to the occlusal area of M1. Both of these features presage the greater degree of molarization seen in extant galagids. The P4 of Progalago, a possible stem galagid from the early Miocene of East Africa, differs in having a longer crown, with a more voluminous talonid basin, and weaker distal cuspules. M1 has four main cusps. The protoconid and metaconid are subequal in height, relatively low, and positioned quite close together. The protoconid is situated slightly more mesially than the metaconid, so that the transverse crest connecting them is slightly oblique. There is a slight trace of a buccal cingulum around the protoconid. The mesial fovea is quite short, with a convex mesial margin. The hypoconid and entoconid are less elevated than the trigonid cusps and are spaced further apart than the protoconid and metaconid. The cristid obliqua, passing mesially from the hypoconid, is long and obliquely directed. The metaconid and entoconid are separated by a shallow lingual notch. The talonid basin is quite broad, but shallow. The distal margin of the tooth is obliquely oriented to the transverse axis of the crown. M2 is subequal in size to M1 and morphologically very similar. It differs in being broader mesially, with the mesial cusps set further apart, and having a more oblique distal margin. M3 is rela-
79
tively narrower than M2, but variable in overall relative size (see Table 5.3). In LAET 78-4702 its occlusal area is slightly larger than that of M2 (110%), whereas in EP 1064/03 it is smaller (86%). The crown narrows distally. It has five cusps, with a variably developed hypoconulid. The hypoconulid heel is well-developed in LAET 78-4702 and relatively weak in LAET 1064/03. The hypoconid and entoconid are reduced in size relative to the trigonid cusps.
Taxonomic and Phylogenetic Relationships Laetolia sadimanensis can be distinguished from all extant galagids by its unique combination of morphological features. These include a relatively deeper and more robust mandibular corpus, a more vertical mandibular symphysis with a greater cross-sectional area and an inverted tear-drop sagittal section, a more robust, lower-crowned, relatively larger and more vertically implanted P2, posterior premolars relatively small in relation to the molars, P4 less molarized, with a shorter and more ovoid crown, a rounded mesial margin, a less prominent metaconid, a shorter and narrower talonid basin, and more weakly developed entoconid and hypoconid, and relatively narrower lower molars. The extent of these difference necessitate including the Laetoli galagid in a separate genus. The distinctive features of the mandible and P2 are best interpreted as autapomorphies (see Walker 1987; Harrison 2010). The hypertrophy and vertical implantation of the P2, the steep inclined and robust symphysis, and the relatively deep corpus are probably functionally linked, and exhibit some degree of convergence on the morphology seen in extant lorisids. However, as noted above, Laetolia sadimanensis appears to be more primitive than all extant galagids in having a less molariform P4. This condition is most closely approximated by Galago spp. among extant galagids, although the latter do have a relatively larger P4 with a more expanded talonid basin. Compared to Miocene galagids, such as Progalago and Komba, Laetolia is more derived in having a greater degree of P4 molarization, with a relatively larger crown, more expansive talonid (compared with Komba) and better-developed distal cuspules (compared with Progalago). Thus, based on this evidence, Laetolia can be inferred to be the sister taxon of all extant galagids (see Fig. 5.2).
Conclusions An additional specimen of a fossil galagid, comprising associated partial right and left mandibular corpora (EP 1064/03), was recovered from the Upper Laetolil Beds at Loc. 10W in 2003.
80
T. Harrison
Laetoli, occurring in riverine and open acacia woodland. Given that all extant galagids are arboreal, and need trees or thorn bushes for feeding and as sleeping sites, we can infer that the vegetation at Laetoli during the Pliocene included at least open woodland or thorn scrub.
Fig. 5.2 Cladogram showing the inferred phylogenetic relationships of Laetolia with Miocene galagids (i.e., Komba and Progalago) and extant galagids (i.e., Galagoides, Otolemur, Sciurocheirus, Galago, Euoticus). Relationships between extant galagids are based on data from Fabre et al. (2009). Morphological features used to place Laetolia as the sister taxon to extant galagids are described in the text
This specimen now represents the most complete specimen of a Pliocene galagid. The material from Laetoli can all be attributed to a single species, previously known as Galago sadimanensis. However, the taxon is considered to be sufficiently distinct from extant and Miocene galagids to be placed in its own genus, Laetolia nov. gen. The fossil record of galagids from the late Miocene and Pliocene is relatively poor, and Laetolia sadimanensis represents the best-known form. Laetolia sadimanensis is distinguished from all extant galagids by its unique combination of features. The stout and vertical implantation of the P2, the steeply inclined and robust symphysis, and the relatively deep corpus are all specialized features and probably functionally linked. However, Laetolia sadimanensis has a less molariform P4 than extant galagids, and it probably represents the primitive sister taxon to crown galagids (see Fig. 5.2). Given that crown galagids are estimated to have shared a last common ancestor during the late Oligocene (Fabre et al. 2009), based on molecular evidence, it is interesting to discover that a rather specialized sister taxon of extant galagids survived in East Africa until at least the midPliocene contemporary with more advanced crown members. Not much can be deduced about the paleoecology at Laetoli based on the rare occurrence of fossil galagids. Modern-day species have a wide distribution throughout sub-Saharan Africa, ranging from tropical forests and dry forests, to acacia woodland, savanna and thorn scrub (Kingdon 1997). Galago senegalensis is found today at
Acknowledgements The author is grateful to the Tanzania Commission for Science and Technology and the Unit of Antiquities in Dar es Salaam for permission to conduct research in Tanzania. Special thanks go to Paul Msemwa (Director) and Amandus Kweka, as well as to all of the staff at the National Museum of Tanzania in Dar es Salaam, for their support and assistance. The Government of Kenya and the National Museums of Kenya are thanked for permission to study the collections in Nairobi. Thanks to Emma Mbua, Mary Muungu, Meave Leakey (Kenya National Museum), Jerry Hooker, Peter Andrews, Paula Jenkins, Daphne Hills (Natural History Museum, London), Nancy Simmons, Ross MacPhee, and Eileen Westwig (American Museum of Natural History, New York) for access to specimens in their care. For their advice, discussion, and help I gratefully acknowledge the following individuals: P. Andrews, E. Delson, C. Jolly, D.M.K. Kamamba, M.G. Leakey, C.S. Msuya, S. Odunga, M. Pickford, L. Pozzi, and D. Su. I am especially grateful to R. Kay and H. Wesselman for their feedback on the manuscript. Research on the Laetoli galagids was supported by grants from the National Geographic Society, the Leakey Foundation, and NSF (grants BCS-9903434 and BCS-0309513).
References Bearder, S. K. (1999). Physical and social diversity among nocturnal primates: A new view based on long term research. Primates, 40, 267–282. Conroy, G. C., Pickford, M., Senut, B., & Mein, P. (1993). Diamonds in the desert: The discovery of Otavipithecus namibiensis. Evolutionary Anthropology, 2, 46–52. Conroy, G. C., Senut, B., Gommery, D., Pickford, M., & Mein, P. (1996). Brief communication: New primate remains from the Miocene of Namibia, southern Africa. American Journal of Physical Anthropology, 99, 487–492. Denys, C. (1987). Micromammals from the West Natron Pleistocene deposits (Tanzania). Biostratigraphy and paleoecology. Sciences Géologiques Bulletin, 40, 185–201. Fabre, P.-H., Rodrigues, A., & Douzery, E. J. P. (2009). Patterns of macroevolution among primates inferred from a supermatrix of mitoichondrial and nuclear DNA. Molecular Phylogenetics and Evolution, 53, 808–825. Gebo, D. L. (1986). Miocene lorisids – the foot evidence. Folia Primatologica, 47, 217–225. Geoffroy Saint-Hilaire, E. (1812). Tableau des quadrumanes, 1. Ord. Quadrumanes. Annales du Musée d’Histoire Naturelle de Paris, 7, 260–273. Gray, J. E. (1821). On the natural arrangement of vertebrose animals. London Medical Repository Record, 15, 296–310. Gray, J. E. (1825). Outline of an attempt at the disposition of the Mammalia into tribes and families with a list of the genera apparently appertaining to each tribe. Annals of Philosophy n.s., 10, 337–344. Gregory, W. K. (1915). On the classification and phylogeny of the Lemuroidea. Bulletin of the Geological Society of America, 26, 426–446. Groves, C. (2001). Primate taxonomy. Washington, DC: Smithsonian Institution Press.
5 Laetoli Galagids Grubb, P., Butynski, T. M., Oates, J. F., Bearder, S. K., Disotell, T. R., Groves, C. P., & Struhsaker, T. T. (2003). Assessment of the diversity of African primates. International Journal of Primatology, 24, 1301–1357. Harris, J. M., Leakey, M. G., & Cerling, T. E. (2003). Early Pliocene tetrapod remains from Kanapoi, Lake Turkana Basin, Kenya. Contributions in Science, 498, 39–113. Harrison, T. (2010). Later Tertiary Lorisiformes. In L. Werdelin & W. J. Sanders (Eds.), Cenozoic mammals of Africa (pp. 333–349). Berkeley: University of California Press. Kingdon, J. (1997). The Kingdon field guide to African mammals. San Diego: Academic. Linnaeus, C. (1758). Systema naturae per regna tria naturae, secundum classes, ordines genera, species cum charateribus, diffentriis, synonymis, locis (10th ed.). Stockholm: Laurentii Salvii. Masters, J. C., & Bragg, N. P. (2000). Morphological correlates of speciation in bush babies. International Journal of Primatology, 21, 793–813. Mein, P., & Pickford, M. (2006). Late Miocene micromammals from the Lukeino Formation (6.1 to 5.8 Ma), Kenya. Bulletin Mensuel de la Société Linnéenne de Lyon, 75, 183–223.
81 Phillips, E., & Walker, A. (2002). Fossil lorisoids. In W. C. Hartwig (Ed.), The primate fossil record (pp. 83–95). Cambridge: Cambridge University Press. Pickford, M., Wanas, H., & Soliman, H. (2006). Indications for a humid climate in the Western Desert of Egypt 11–10 Myr ago: Evidence from Galagidae (Primates, Mammalia). Comptes Rendus Palevol, 5, 935–943. Rasmussen, D. T., & Nekaris, K. A. (1998). Evolutionary history of lorisiform primates. Folia Primatologica, 69 (Suppl. 1), 250–285. Simpson, G. G. (1965). Family: Galagidae. In L. S. B. Leakey (Ed.), Olduvai Gorge 1951–61: A preliminary report on the geology and fauna (Vol. 1, pp. 15–16). Cambridge: Cambridge University Press. Szalay, F. S., & Delson, E. (1979). Evolutionary history of the primates. New York: Academic. Walker, A. C. (1987). Fossil Galaginae from Laetoli. In M. D. Leakey & J. M. Harris (Eds.), Laetoli: A Pliocene site in northern Tanzania (pp. 88–90). Oxford: Clarendon. Wesselman, H. B. (1984). The Omo micromammals: Systematics and paleoecology of early man sites from Ethiopia (Contributions to vertebrate evolution, Vol. 7). Basel: Karger.
wwwwwwwwwwwwwww
Chapter 6
Cercopithecids (Cercopithecidae, Primates) Terry Harrison
Abstract New finds from Laetoli have allowed a more detailed assessment of the taxonomy and paleobiology of the fossil cercopithecids. Most of the specimens consist of isolated teeth, jaw fragments and postcranial bones from the Upper Laetolil Beds (~3.5–3.8 Ma), but four specimens are known from the Upper Ndolanya Beds (~2.66 Ma) and a proximal humerus has been recovered from the Lower Laetolil Beds (~3.8–4.3 Ma). Four species are represented: Parapapio ado, Papionini gen. et sp. indet., cf. Rhinocolobus sp., and Cercopithecoides sp. Parapapio ado is the most common species. Based on dental size and proportions and facial morphology, Pp. ado can be distinguished from all other species of Parapapio. The postcranial specimens attributed to Pp. ado indicate that it was a slender and agile semi-terrestrial monkey. A few isolated teeth represent a second species of papionin, larger in dental size than Pp. ado. Due to the paucity of the material, the taxon is left unassigned at the genus and species level. A distal humerus attributed to this taxon indicates that it was large terrestrial cercopithecid. The most common species of colobine is referred to cf. Rhinocolobus sp., based on its overall similarities to Rhinocolobus turkanaensis. The material can be distinguished from all fossil colobine species previously recognized from Africa, but without more complete cranial specimens it is not possible to diagnose a new taxon. From the postcranial material it can be inferred that it was generally adapted for arboreal quadrupedalism. The somewhat smaller species of colobine represents a previously undescribed species of Cercopithecoides. The postcranial specimens attributed to this taxon indicate that it was fully arboreal. Analysis of the distribution of the Laetoli cercopithecids provides provisional evidence of spatial patterning and temporal trends. For example, the dentition of Parapapio exhibits a trend to increase in size during the course of the Upper Laetolil Beds. As at other late Miocene and early Pliocene localities older than 3.5 Ma, the Laetoli cercopithecid community is characterized by the absence of Theropithecus and the relatively T. Harrison (*) Center for the Study of Human Origins, Department of Anthropology, New York University, 25 Waverly Place, New York, NY 10003, USA e-mail:
[email protected] large proportion of colobines. After 3.5 Ma Theropithecus becomes the dominant cercopithecid at all East African localities, and the proportion of colobines declines accordingly. Keywords Parapapio • Paracolobus • Rhinocolobus • Cercopithecoides • Papionin • Colobines • Monkey • Pliocene • East Africa
Introduction Fossil cercopithecids were first discovered at Laetoli by L.S.B. Leakey in 1935. These included two mandibular fragments of a small to medium-sized species of papionin that were forwarded to the Natural History Museum in London. Hopwood (1936) described the right mandibular corpus of a female individual (NHM M14940) from the 1935 collection, and made it the holotype of a new species, Cercocebus ado. In 1938–1939 Kohl-Larsen made extensive collections of fossil vertebrates in the Laetoli region, and these included 38 cercopithecids, now housed in the Humboldt-Universität Museum für Naturkunde in Berlin. Although precise locality information is lacking, most of the specimens were recorded as coming from the Garussi and Vogelfluss (= Garusi), Deturi Ost (= Olaitoli) and Marambu (= Locality 1) valleys (see Harrison and Kweka 2011), and presumably all of them derive from the Upper Laetolil Beds. The preservation of the fossils and lithology of the adhering matrix support such a provenience. Two additional cercopithecid specimens from the Kohl-Larsen collection, attributable to Papio sp., were recovered from a Pleistocene locality called Lemagrut Korongo on the northwestern slope of Lemagurut. Dietrich (1942) published a brief account of the cercopithecids from the Kohl-Larsen collection. He erected a new species Papio (Simopithecus) serengetensis for the mediumsized papionin, assuming that the assigned material was distinct from Cercocebus ado Hopwood, 1936. Leakey and Delson (1987) presumed that Dietrich’s “kurzschnauzigen Catarrhinen” related to short-faced colobines from Laetoli,
T. Harrison (ed.), Paleontology and Geology of Laetoli: Human Evolution in Context. Volume 2: Fossil Hominins and the Associated Fauna, Vertebrate Paleobiology and Paleoanthropology, DOI 10.1007/978-90-481-9962-4_6, © Springer Science+Business Media B.V. 2011
83
84
although Dietrich was evidently referring to Cercocebus ado (Dietrich 1942: 53). They also mistakenly indicated that Reck and Kohl-Larsen (1936) reported cercopithecids from Laetoli (Cercocebus sp. and Papio sp.), when, in fact, this material was from the Pleistocene locality of Eyasi. L. S. B. Leakey and M. D. Leakey briefly revisited Laetoli in 1959 and 1964, and recovered additional fossil cercopithecids. According to M. D. Leakey (1987a), these finds and the 1935 collections were mainly from the Locality 10 complex (Locs. 10, 10W and 10E). Leakey and Leakey (1976) assigned the specimens to Cercocebus ado. Delson (1978; Szalay and Delson 1979), based on his study of the specimens in Berlin, recognized? Parapapio ado and Colobinae gen. et sp. indet. in the Laetoli collections. Mary Leakey’s expeditions recovered 81 cranio-dental specimens of fossil cercopithecids at Laetoli from 1974 to 1979. Locality information is recorded for most of these specimens, but precise stratigraphic information is largely unknown. All of the specimens come from the Upper Laetolil Beds, with the exception of two isolated teeth from the Upper Ndolanya Beds (Loc. 7E). A detailed systematic account of the entire Laetoli collection was presented by Leakey and Delson (1987). They recognized four species of cercopithecids – Parapapio ado, cf. Papio sp., cf. Paracolobus sp., and Colobinae gen. et sp. indet. The fragmentary nature of the material prevented more precise taxonomic assignments. They designated a lectotype (MB Ma 42441 = MB 1938.1) for Papio (Simopithecus) serengetensis Dietrich, 1942, and recognized the nomen as a junior synonym of Parapapio ado (Hopwood 1936). The present author has recovered an additional 83 craniodental specimens of cercopithecids in the course of his 1998–2005 fieldwork at Laetoli (Table 6.1). These are all from the Upper Laetolil Beds, except for a deciduous upper central incisor of Parapapio from the Upper Ndolanya Beds at Loc. 7E. In addition, a proximal humerus from Emboremony 1 represents the first cercopithecid recovered from the Lower Laetolil Beds. The new material from Laetoli consists primarily of isolated teeth and jaw fragments, so several of the long-standing taxonomic issues remain unresolved. However, the larger sample of specimens now available (more than 200 cranio-dental specimens; see Table 6.1), the discovery of several key finds, and a detailed reassessment of the morphology has helped to clarify the taxonomic relationships of the Laetoli cercopithecids, as well as improving our understanding of their paleobiology. Following Leakey and Delson (1987), four cercopithecid species are recognized – two papionins and two colobines. These are recognized here as Parapapio ado, Papionini gen. et sp. indet., cf. Rhinocolobus sp., and Cercopithecoides sp. Few of the postcranial remains are directly associated with cranio-dental specimens, but most can be assigned, at least provisionally, to a specific taxon on their basis of size and
T. Harrison Table 6.1 Distribution by locality of cercopithecid cranio-dental specimens from Laetoli Papionini cf. Parapapio Rhinocolobus Cercopithe- gen. et sp. indet. sp. coides sp. Main locality ado Laetolil Beds 1 2 3 4 5 6 7 8 9 9S 10 10W 10E 11 12 + 12E 13 15 16 17 19 20 21 22 22E
4 12 5 0 1 8 11 12 6 0 3 1 17 9 0 2 0 6 1 0 1 5 4 0
2 4 4 0 0 3 2 2 7 1 1 0 1 5 2 1 0 1 2 0 0 4 0 0
0 0 0 0 0 1 1 0 0 0 0 0 2 1 0 0 0 0 0 0 0 1 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 1 0 0 0 0 0 0 0 0
Ndolanya Beds 7E 2 14 0 15 0 18 0 Silal Artum 0 Unknown 29 Total 139
1 0 0 0 0 18 61
0 0 0 0 0 3 9
0 0 0 0 0 1 3
morphology. The aim of this chapter is to present a detailed description and comparison of the cranio-dental morphology of each species, and a preliminary account of the postcranial material. This provides the basis for a reassessment of the taxonomy of the Laetoli cercopithecids, as well as some initial observations on their paleobiology and ecology.
Material and Methods The sample of cercopithecids from Laetoli comprises 212 cranio-dental specimens and 25 postcranial specimens (Table 6.1). This includes 93 specimens recovered from 1998 to 2005 and described here for the first time. Almost all of the cercopithecids from Laetoli have been recovered from the Upper Laetolil Beds. The exceptions are three isolated teeth from the Upper Ndolanya Beds at 7E and a proximal
6 Laetoli Cercopithecids
humerus from the Lower Laetolil Beds at Emboremony 1. A catalog of fossil cercopithecids from Laetoli is presented in Tables 6.2, 6.5, 6.6, 6.9 and 6.11. The cranio-dental specimens consist primarily of isolated teeth (62%). Although a number of cranial and mandibular specimens are represented, these are rather fragmentary and there are no partial or complete crania and mandibles, which hampers comparisons and taxonomic assessments. The relative paucity of postcranial remains is not due to a collecting bias, but rather is a consequence of the taphonomic impact that carnivore predation and scavenging had on the composition of the fossil assemblage (Su and Harrison 2008). Cercopithecids, like the hominins, show a markedly disproportionate representation of cranio-dental specimens over postcranial specimens. Disarticulation of the skeleton and much of the damage to individual bones occurred prior to burial and fossilization, but additional damage was caused after the specimens eroded out of the sediments by weathering and transportation, and especially by trampling (Su and Harrison 2008). Although none of the cranial remains (and few of the postcranials) shows evidence of carnivore bite marks, 11 (6.7%) of the teeth collected by Leakey and Harrison show signs of having been digested (Carter-Menn unpublished data), confirming that carnivores had an important taphonomic impact on the cercopithecid assemblage. The Laetoli specimens described here are housed in the Natural History Museum in London (NHM.M; 1935 Leakey collection), Humboldt-Universität Museum für Naturkunde in Berlin (MB Ma.; 1938–1939 Kohl-Larsen collection), Kenya National Museum in Nairobi (LIT and LAET; 1959, 1964 and 1974–1979 Leakey collections on loan from Tanzania), and National Museum of Tanzania (EP, Eyasi Plateau expedition; 1998–2005 Harrison collection). Comparison with extant and fossil cercopithecids were carried out at the American Museum of Natural History (AMNH), Natural History Museum in London (NHM), and Kenya National Museum (KNM). Molar terminology follows Jolly (1972) and Delson (1975), with additional crest terminology following Szalay and Delson (1979). Metrical data on extant primates were collected by the author, and supplemented by dental metrics from Swindler (2002).
Systematics and Description Order Primates Linnaeus, 1758 Infraorder Catarrhini Geoffroy, 1812 Superfamily Cercopithecoidea Gray, 1821 Family Cercopithecidae Gray, 1821 Subfamily Cercopithecinae Gray, 1821 Tribe Papionini Burnett, 1828
85
Parapapio Jones, 1937 Diagnosis: Cranium characterized by lack of an anteorbital drop (i.e., line from glabella to nasion straight or gently concave). Supraorbital tori relatively thin, and do not project anteriorly. Supraorbital sulcus (= ophryonic groove) weakly developed or absent. Suborbital fossae and well-developed maxillary ridges are generally absent, although a shallow depression may be present in some individuals. Fossae on lateral side of mandibular corpus weakly excavated or absent. Postcranial morphology indicates more arboreal positional behaviors than Papio and Theropithecus (Adapted from Freedman 1957; Szalay and Delson 1979; Leakey and Delson 1987; Frost and Delson 2002; Heaton 2006). Distribution and Taxonomy Five species of Parapapio are currently recognized (Jablonski 2002; Gilbert 2007; Frost 2007): Pp. ado Hopwood, 1936; Pp. broomi Jones, 1937; Pp. jonesi Broom, 1940; Pp. whitei Broom, 1940; Pp. lothagamensis Leakey et al., 2003. The alpha-taxonomy and assignment of specimens to Parapapio species from Plio-Pleistocene localities in South Africa have proved problematic (Eisenhart 1974; Freedman 1957, 1976; Szalay and Delson 1979; Frost and Delson 2002; Jablonski 2002; Heaton 2006; Williams et al. 2007; Gilbert 2007, 2008). Most researchers currently recognize three species from South Africa, distinguished primarily on the basis of size, as well as aspects of the facial morphology (Jablonski 2002; Frost and Delson 2002; El-Zaatari et al. 2005; Frost 2007). Parapapio jonesi is the smallest species, followed in size progression by Pp. broomi and Pp. whitei. All three species occur contemporaneously at Sterkfontein and Makapansgat (~3.3–2.3 Ma), while Pp. broomi (Taung and Bolt’s Farm) and possibly Pp. jonesi (Taung and Kromdaai A) extend their temporal range to ~2.3–2.0 Ma and ~1.5–1.0 Ma respectively (Jablonski 2002; El-Zaatari et al. 2005). Similar material has been recovered from Plio-Pleistocene cave sites on the Humpata Plateau in Angola and in the Koanaka Hills in Botswana, but they have not yet been attributed to species (Pickford et al. 1992; Jablonski 1994, 2002; Senut 1996). Fragmentary remains of Parapapio sp. of midto late Pliocene age have also been recovered from the Chiwondo Beds of Malawi (Bromage and Schrenk 1986; Bromage et al. 1995; Frost and Kullmer 2008). Two isolated teeth from the Quartzose Sand Member of the Varswater Formation (Fm.) at Langebaanweg, South Africa (~5.0 Ma), also have their closest affinities with teeth of Parapapio (Grine and Hendey 1981). Two of the South African species have been recorded provisionally from localities in East Africa. Parapapio cf. jonesi is identified in the Hadar Fm. (~3.4–2.9 Ma) in Ethiopia
86
(Frost and Delson 2002), and Parapapio cf. Pp. whitei is recorded from the Lomekwi Mb., Nachukui Fm., West Turkana (~2.5–3.3 Ma) in northern Kenya (Harris et al. 1988). Parapapio ado, which is slightly larger in mean dental size than P. jonesi, is known definitively only from Laetoli, the type locality (~3.5–3.8 Ma). A few isolated teeth of this species are recorded from the younger Upper Ndolanya Beds at Laetoli (Leakey and Delson 1987; see below), and these extend the temporal range of the taxon to ~2.6–2.7 Ma. Patterson (1968) referred a partial mandible from Kanapoi (~4.1–4.2 Ma) to Pp. jonesi, but with the recovery of a relatively large sample of cercopithecid specimens from this site in the 1990s (Leakey et al. 1995, 2003), the mandible fragment and other Parapapio specimens were attributed to Pp. ado. Parapapio ado has also been provisionally identified from the lower Lomekwi Mb. of the Nachukui Fm. (~3.4 Ma) in West Turkana (Area 106) and in the Koobi Fora Fm. (~4.0–3.4 Ma) in East Turkana, Kenya (Harris et al. 1988; Leakey et al. 1995; Jablonski et al. 2008a). A new species of Parapapio, Pp. lothagamensis, was described by Leakey et al. (2003) from the Upper and Lower Nawata Fm. of Lothagam, Kenya (~5.0–7.4 Ma). It can be distinguished from other species of Parapapio by its smaller size and by a suite of distinctive features of its dentition and mandible. Unidentified species of Parapapio or other small papionins are recorded from the Omo Shungura Mb. B-lower G and the lower part of the Nachukui and Koobi Fora Fms. in the Lake Turkana Basin of Ethiopia and Kenya (Leakey and Leakey 1976; Eck 1976, 1977; Delson 1984; Jablonski et al. 2008a), the Nkondo Fm. in Uganda (~3.6 Ma) (Senut 1994), Unit 2 of the Chiwondo Beds (~4 Ma or older) in Malawi (Frost and Kullmer 2008), and the late Miocene (~6–7 Ma) of As Sahabi in Libya (Benefit et al. 2008). Jablonski et al. (2008a) have recently recognized three additional unnamed morphs of Parapapio from the Koobi Fora Fm. – sp. indet. A (2.0–1.4 Ma), sp. indet. B (4.0–1.6 Ma), and sp. indet. C (3.6–1.4 Ma). In addition to Parapapio, two genera of small- to mediumsized papionins have recently been diagnosed and described from East and South Africa respectively. Frost (2001) described a new species of papionin, Pliopapio alemui, from Aramis, Ethiopia (~4.4 Ma), which differs in facial morphology and overall size from Parapapio. Additional specimens tentatively attributed to this species from the Sagantole and Adu-Asa Formations in the Middle Awash region of Ethiopia may extend the taxon back to ~5.7 Ma (Haile-Selassie et al. 2004; Frost et al. 2009). Gilbert (2007) transferred Parapapio antiquus (Haughton 1925) from Taung (~2.3–2.0 Ma) to a new genus, Procercocebus, based on its inferred phylogenetic relationship with the Cercocebus + Mandrillus clade. Finally, Jablonski et al. (2008a) identified a small species of papionin from Koobi Fora (~1.9–1.4 Ma) as Lophocebus cf. albigena.
T. Harrison
Parapapio ado (Hopwood, 1936) This is the most common species of cercopithecid from Laetoli, representing 65.6% of the cranio-dental specimens (Table 6.2). All of the permanent teeth are represented in the collections, with the exception of I2 (see Table 6.3 for dimensions). A number of mandibular specimens and cranial fragments are represented, but the lack of relatively complete skulls limits comparisons with other extant and fossil papionins. The most important new finds consist of a mandible of a large male individual with almost complete, but heavily worn, dentition (EP 700/00); a partial frontal bone (EP 1579/98), which, when combined with information from LAET 75-2966 (a previously unattributed frontal found by M.D. Leakey) provides the first evidence of the morphology of the upper face of Pp. ado; and a right maxilla of a juvenile individual with dP3-dP4 and M1 (EP 900/03). Several postcranial remains (n = 10) are attributed to Parapapio ado, and they are represented proportionately in the Laetoli collections relative to the cranio-dental remains (see Table 6.11). A brief account of the morphology and functional/behavioral implications of the postcranial specimens is presented later in this chapter. All but two of the specimens of known provenance were recovered from the Upper Laetolil Beds. The two isolated teeth from the Upper Ndolanya Beds at Loc. 7E (LAET 79-5472, M3; EP 1215/03, dI1) are morphologically and metrically indistinguishable from those from the upper Laetolil Beds, and are referred to the same species. When the samples of Parapapio from different stratigraphic units within the Upper Laetolil Beds are considered, there are significant differences in size (but apparently not morphology) between samples earlier and later in the sequence (i.e., below and above Tuff 5 respectively). The mesiodistal lengths of the upper and lower molars from below Tuff 5 (n = 10) are significantly smaller than those from above Tuff 5 (n = 67) (expressed as a standard deviation from the mean for each tooth type for the sample above Tuff 5; Student’s t-Test, t = 2.03, df = 75, p 72.5 >185.0 Kuseracolobus aramisi Aramis 49.0 64.2 85.0 198.2 Cercopithecoides meavae Hadar & Leadu 57.3 66.3 86.6 210.2 Cercopithecoides sp. Laetoli 52.7 68.3 95.0 216.0 Cercopithecoides williamsi Sterkfontein 57.4 72.8 95.9 226.1 Cercopithecoides williamsi East Turkana 62.5 76.3 111.0 249.8 Kuseracolobus hafu Asa Issie 72.0 90.5 115.6 278.1 Rhinocolobus turkanaensis East Turkana & Omo 69.6 85.6 123.8 279.0 cf. Rhinocolobus turkanaensis Hadar 66.8 93.5 123.1 283.4 Cercopithecoides kimeui Koobi Fora 81.7 98.0 111.9 291.6 cf. Rhinocolobus sp. Laetoli 72.5 96.8 129.2 298.5 Paracolobus chemeroni Chemeron 94.9 125.0 158.4 378.3 Paracolobus mutiwa Omo 84.2 115.9 158.7 358.8 Data: Freedman (1957), M.G. Leakey (1982, 1987c), Frost (2001), Frost and Delson (2002), Leakey et al. (2003), Hlusko (2006, 2007), Jablonski et al. (2008b); Harrison, unpublished data
a perture, anterior root of the zygomatic process situated opposite M1-M2, laterally curved cheek tooth row, and a relatively short palate, and in these respects it matches well with R. turkanaensis, Kuseracolobus, and Cercopithecoides. It differs from Rhinocolobus turkanaensis in having a shallower subnasal clivus, a less robust frontal process of the zygoma, and a shorter rostrum. Paracolobus mutiwa and Pc. chemeroni differ in having a U-shaped inferior margin to the nasal aperture, a relatively longer palate, a more posteriorly placed anterior root to the zygomatic arch (opposite M3 in Pc. mutiwa and M2 in Pc. chemeroni), relatively straight-sided tooth rows, and a broader muzzle with distinct facial fossae. The large maxillary sinus present in the Laetoli specimens also occurs in Cercopithecoides williamsi from South Africa (absent in specimens from the Turkana Basin), C. kimeui, and possibly C. alemayehui, as well as Libypithecus and Mesopithecus, but is absent in Rhinocolobus and Paracolobus (Rae 2008; Gilbert and Frost 2008; Harrison, unpublished observation). The mandible of cf. Rhinocolobus sp. from Laetoli resembles Rhinocolobus turkanaensis in having a relatively long and sloping symphysis, no median mental foramen (absent in R. turkanaensis from Hadar, but present in the Turkana Basin material), a corpus that deepens posteriorly, a narrow extramolar sulcus, and absence of prominentia laterales. It differs from Paracolobus and Kuseracolobus in having a less vertical symphysis, a corpus that does not deepen posteriorly to such a marked extent (similar to Pc. enkorikae), and absence of prominentia laterales. The Laetoli large colobine differs from Cercopithecoides in having a more sloping symphysis with a long subincisive planum, absence of a median mental foramen (present in C. williamsi and C. kimeui), a deeper and more slender corpus that deepens slightly posteriorly, and absence of prominentia laterales (except C. meavae).
In sum, the large colobine from Laetoli bears its closest similarity in its dentition and lower face to Rhinocolobus turkanaensis. It differs mainly in having a shorter rostrum, a shallower subnasal clivus, a more gracile frontal process of the maxilla, a maxillary sinus, a slightly shallower mandibular corpus, and relatively shorter upper and lower molars. Based on these comparisons, the taxon represented at Laetoli is tentatively retained in Rhinocolobus, although it clearly represents a distinct species from R. turkanaensis. With the recovery of more complete material, it is possible that the Laetoli specimens could belong to a novel genus. While my own comparisons (see below) agree with the observations of Leakey and Delson (1987) that the proximal femur corresponds closely in morphology to that of Pc. chemeroni (although it differs from that of Pc. mutiwa), without comparative material of Rhinocolobus turkanaensis, it is not possible to conclude that the proximal femur from Laetoli provides grounds for attribution to any particular genus of colobine.
Cercopithecoides Mollett, 1947 Diagnosis: Medium-sized to very large colobines with globular calvaria. Rostrum short in relation to neurocranial length. Frontal process of zygomatic bone narrow. Interorbital region broad. Supraorbital tori thick, with deep supraorbital sulcus. Sagittal crest absent, at least anteriorly. Mandibular symphysis steep, but shallow with a median mental foramen (except C. meavae). Mandibular corpus shallow and thick, with slightly expanded or unexpanded gonial region. P3 protocone reduced (Adapted from Freedman 1957; Leakey 1982; Szalay and Delson 1979; Frost and Delson 2002).
118
Taxonomy and Distribution Five species of Cercopithecoides are recognized – C. williamsi Mollett, 1947; C. kimeui Leakey, 1982; C. meavae Frost and Delson, 2002; C. kerioensis Leakey et al., 2003; and C. alemayehui Gilbert and Frost, 2008. Cercopithecoides williamsi is the only large colobine represented at PlioPleistocene localities (~3.2–1.0 Ma) in South Africa (Freedman 1957; Jablonski 2002). This same species occurs contemporaneously at Koobi Fora (~3.4–1.5 Ma) in Kenya, but is much less common. Cercopithecoides kimeui is known from Koobi Fora, West Turkana, Hadar (Pinnacle Locality), Olduvai Beds II-III, and Rawi in eastern Africa, dating from ~3.4–0.8 Ma (Leakey 1982; Frost and Delson 2002; Jablonski 2002; Frost et al. 2003; Frost 2007; Jablonski et al. 2008b). A third species, C. meavae from Leadu and Hadar (~3.4– 3.3 Ma), occurs contemporaneously in eastern Africa. A small species of Cercopithecoides, C. kerioensis, from Lothagam (possibly from the Apak Member, ~4.2–5.0 Ma), may represent the earliest representative of the genus (Leakey et al. 2003). Finally, a recently named species, C. alemayehui (Gilbert and Frost 2008), known only from the type specimen from the Daka Member, Bouri Formation in the Middle Awash region of Ethiopia (~1.0 Ma), is apparently a late surviving species in eastern Africa. As indicated in the diagnosis, all of these species are characterized by a distinctive suite of specialized cranio-dental features, including a short rostrum, a steep mandibular symphysis, a shallow and robust mandibular corpus that maintains a constant depth below the cheek teeth, a slightly expanded gonial region, and a P3 with a reduced protocone (Frost and Delson 2002; Jablonski 2002; Jablonski et al. 2008b). In addition, all of the known species for which postcranials are known (not C. kerioensis or C. alemayehui) are apparently specialized for terrestrial locomotion (Frost and Delson 2002; Jablonski et al. 2008b).
Cercopithecoides sp. Material assigned to this genus from Laetoli was first attributed to Colobinae gen. et sp. indet. by Leakey and Delson (1987). The sample consisted of two upper premolars from the Kohl-Larsen collection and two upper premolars from the Leakey collection, as well as a cuboid and a proximal femur assigned on the basis of size. The teeth were recognized as smaller than those assigned to cf. Rhinocolobus sp. Although only a few additional specimens of the small colobine from Laetoli have been recovered since 1998 (see Table 6.9), the sample now includes a partial mandible with P3-M3 of a female individual (EP 1425/04), which helps to resolve its taxonomic affinities (Fig. 6.16). The mandibular corpus in EP
T. Harrison Table 6.9 List of cranio-dental specimens from Laetoli referred to Cercopithecoides sp. Specimena Loc.b Horizonc Element and commentsd MB Ma 42477 MB Ma 42478 LAET 75-3372a LAET 77-4565 LAET 77-4578 EP 201/98 EP 202/98 EP 1079/04 EP 1425/04
Rt P4. [MB 1939.16.20] Rt P3. [MB 1939.16.27] 21 Rt P3 ? Rt M3 7 Lt P4 10E Tuffs 5-7 Rt M3 10E Tuffs 5-7 Rt M1 11 Tuffs 7-8 Rt M3 6 Tuffs 5-7 Lt mandibular fragment with P3-M3. Female a Specimen prefixes: MB Ma, Humboldt-Universität Museum für Naturkunde, Berlin, 1938–1939 Ludwig Kohl-Larsen collection; LAET, Kenya National Museum, Nairobi (on loan from Tanzanian National Museum), 1974–1979 Mary Leakey collections; EP, Eyasi Plateau Expedition, National Museum of Tanzania, Dar es Salaam, 1998–2005 Terry Harrison collections b Localities: Numbers refer to the collecting localities designated by Leakey (1987a) c Horizon: The stratigraphic provenience of the Kohl-Larsen collections and most of the specimens collected by Leakey are from unknown horizons within the Upper Laetolil Beds, except where indicated. The Harrison collections are mostly surface finds, and the stratigraphic provenience is recorded as a fossiliferous section between two marker tuffs within the Upper Laetolil Beds (unless more precise provenience is known for in situ specimens). All specimens are from the Upper Laetolil Beds d Element and comments: lt, left; rt, right. Sex is determined by the size and morphology of the canines and P3. The Museum für Naturkunde in Berlin has recently provided new accession numbers for their fossil mammal collections; the previous numbers, listed by Leakey and Delson (1987), are cross-referenced here
1425/04 is low and robust, and does not increase in depth posteriorly, with a slightly convex inferior margin. In addition, the P3 has a very reduced protocone, and the lower molars exhibit a very marked wear differential between the lingual and buccal cusps. This combination of morphological features is typical of Cercopithecoides, and serves to distinguish it from all other Plio-Pleistocene colobines. Attribution of the smaller colobine from Laetoli to Cercopithecoides seems well justified. However, the Laetoli material has a unique set of characters that distinguish it from all of the currently recognized species (see comparisons below). It certainly represents a different species of Cercopithecoides, but unfortunately the current material is not adequate to diagnose a new taxon. It is recognized here as Cercopithecoides sp.
Description of Cranio-Dental Morphology No cranial specimens of this taxon have been recovered from Laetoli. The mandibular corpus, known only from EP 1425/04, is quite low and robust (see Fig. 6.16). Mandibular robusticity at M2 is 50.8 (mean breadth-height index of corpus at M2).
6 Laetoli Cercopithecids
Fig. 6.16 Cercopithecoides sp. EP 1425/04, right mandibular corpus with P3–M3. (a) lateral view, (b) occlusal view, (c) medial view
It shallows slightly posteriorly below the cheek teeth. Laterally, the corpus is slightly concave below P3-M1, and dorsoventrally convex below M2-M3. There is no mandibular corpus fossa. The root of the ramus originates below M3, and judging from the preserved portion it ascended just posterior to M3 without overlap in lateral view. It lacks a distinct prominentia laterales. A single large elliptical mental foramen occurs below P3/P4, about two-thirds down the corpus from the alveolar margin. The inferior margin of the corpus is quite thick and well rounded, with a slightly convex anteroposterior contour. It curves medially as it passes posteriorly. The symphyseal region is incompletely preserved, with the inferior margin missing. There is a low, rounded superior transverse torus, separated from the simian shelf by a welldefined genioglossal pit. A low rounded ridge, the mylohyoid line, originates from the simian shelf and passes posteriorly and superiorly along the medial face of the corpus, and eventually fades midway below M3. The area superior to the mylohyoid line is generally convex. Below the line is a
119
deeply concave area, at least anteriorly, representing the submandibular fossa. The posterior wall of the alveolus for the lower canine is preserved anterior to P3. It is evident that the root was short and slender, characteristic of female individuals, and that it abutted directly against the root of P3, without a diastema. The cheek tooth row shows a slight lateral curvature. Examples of the dentition are limited to the upper premolars and the lower cheek teeth (P3-M3) (Fig. 6.17, Table 6.9). Measurements of the dentition are presented in Table 6.10. P3 is a narrow, subtriangular tooth with two cusps that are markedly different in elevation. The paracone is very tall, and cristodont, while the protocone is reduced to a tiny conical tubercle. The pre- and postparacrista are long and sharp, and subequal in length. The mesial and distal crests of the protocone are short and rounded. A transverse crest descends from the apex of the paracone and passes mesiolingually to terminate midway along the mesial marginal ridge. It does not reach the protocone. As a consequence, a deep longitudinal valley separates the cusps, and the mesial fovea is restricted to a small triangular basin along the mesial margin of the paracone. The distal fovea is expansive. The lingual face of the crown is narrow and strongly convex. The buccal face is remarkably tall, and mesiodistally convex, except for a shallow apico-basal groove mesially. The enamel margin extends further down on the mesiobuccal root than the distobuccal root. P4 shows the same suite of distinctive traits as P3. The crown is ovoid to sub-triangular in shape with two main cusps. The paracone is well-developed, very tall, with sharp pre- and postparacrista. The protocone is weakly expressed, forming a small triangular tubercle on the lingual margin. A fine transverse crest descends from the apex of the paracone and passes lingually and slightly mesially to terminate at the base of the preprotocrista. It does not meet the protocone. The mesial fovea is a small, shallow triangular basin. In both upper premolars there are three roots that are partially or fully fused (unlike the separate roots seen in cf. Rhinocolobus sp.). P3 is relatively small, with a short mesiobuccal honing face for occlusion with the upper canine, at least in female individuals. The enamel junction of the honing face extends only a short distance onto the mesial root. The crown is relatively narrow, and the long-axis is aligned with the posterior cheek teeth. The protoconid is relatively low, and positioned almost centrally on the occlusal surface. The mesial crest is short and sharp. The lingual face is triangular and slightly convex, and is bordered by an indistinct lingual cingulum. The distolingual crest is low and rounded, and terminates basally at the lingual cingulum. The distal basin is a narrow, relatively small and shallow depression, bordered by a low distal marginal ridge. P4 is relatively long and narrow, with its long-axis aligned with the lower molar row. The protoconid and paraconid are
120
T. Harrison
Fig. 6.17 Cercopithecoides sp. (a–c) EP 201/98, right M3. (a) buccal view, (b) lingual view, (c) occlusal view. (d–e) EP 1079/04, right M3. (d) buccal view, (e) occlusal view. (f, g) EP 202/98, right M1. (f) lingual view, (g) occlusal view
transversely aligned, and connected by a well-developed crest. The protoconid is slightly more elevated than the paraconid. The mesial fovea is very short, so the cusps are positioned far mesially. There is a strong preprotocristid, but no preparacristid, so the mesial fovea opens lingually. The talonid basin is long and narrow. Crests descend distally from the apices of the two main cusps to meet the distal marginal ridge. There are no distal tubercles. M1 is a long and narrow tooth (the mean breadth-length index is 80.2), with relatively pronounced buccolingual waisting (see Figs. 6.16 and 6.17f–g). The lingual cusps are quite low for colobines, and are separated by a moderately deep V-shaped lingual notch that extends about halfway down the crown. The buccal cusps are lower, and exhibit a stronger wear differential than the lingual cusps. The protolophid and hypolophid are sharp, elevated and well-developed. The latter is slightly obliquely oriented relative to the transverse axis of the crown. The median buccal cleft forms a well-defined triangular platform between the protoconid
and hypoconid. The mesial and distal buccal clefts are weakly expressed. The buccal marginal crest linking the protoconid and hypoconid is relatively low and rounded, but well-developed. The mesial and distal foveae are relatively small. The talonid basin is rectangular, being longer than broad, and relatively deep. The floor of the basin is transected by an ill-defined Y-shaped groove system, in which the main arm passes through the lingual notch and subsidiary arms pass towards the protoconid and hypoconid. Otherwise, the thin enamel covering is smooth and devoid of secondary wrinkling. M2 is much larger in size than M1, with a marked size differential between them, but they are similar in overall proportions (the breadth-length index is 82.4). M2 differs from M1 in having less marked buccolingual waisting of the crown, broader and more expansive mesial and distal foveae, a deeper talonid basin, more elevated cusps and a deeper lingual notch that extends below mid-crown height. M3 is much longer and slightly broader than M2 (with a mean breadth-length index of 60.8), with only a slight degree
6 Laetoli Cercopithecids
121
Table 6.10 Measurements of teeth of Cercopithecoides sp. from Laetoli Tooth Dimension N Mean SD Range UPPERS P3
P4
MD BL BHT LHT MD BL BHT LHT
3 3 3 3 1 1 1 1
5.9 7.2 8.1 4.8 5.0 6.2 6.8 5.1
0.19 0.54 1.07 0.46 – – – –
5.8–6.2 6.6–7.9 6.9–9.5 4.3–5.4 5.0 6.2 6.8 5.1
LOWERS P3
MD 1 9.5 – 9.5 BL 1 5.2 – 5.2 BHT 1 6.1 – 6.1 HHT 1 9.2 – 9.2 P4 MD 1 7.8 – 7.8 BL 1 6.1 – 6.1 M1 MD 2 8.1 – 7.9–8.2 BLmes 1 6.3 – 6.3 BLdist 2 6.5 – 6.3–6.6 M2 MD 1 9.1 – 9.1 BLmes 1 7.5 – 7.5 BLdist 1 7.2 – 7.2 M3 MD 4 12.5 0.38 12.1–13.0 BLmes 3 7.6 0.31 7.3–8.0 BLdist 4 7.4 0.46 6.8–8.0 BHT buccal height of crown, BL buccolingual breadth, BLmes buccolingual breadth mesially, BLdist buccolingual breadth distally, LHT lingual height of crown HHT length of mesiobuccal face of P3, MD mesiodistal length N number of specimens, SD standard deviation
of buccolingual waisting of the crown (see Figs. 6.16 and 6.17a–e). The lingual cusps are moderately tall, and the lingual notch extends well below mid-crown height. The protolophid and hypolophid are well-developed, being elevated and sharp. The lophids are subparallel, and oriented distinctly obliquely to the transverse axis of the tooth. The mesial fovea is short, broad and very shallow, being restricted to a small crescentic fissure. The talonid basin is deep and rectangular in shape, being slightly longer than broad. It has a simple Y-shaped fissure pattern as in the other lower molars. The median buccal cleft is quite restricted compared with M2, and forms an elliptical pit. Distally there is a short and relatively narrow heel, dominated by a low rounded, but voluminous hypoconulid. The hypoconulid is positioned to the buccal side of the midline, almost in line with the protoconid and hypoconid. It is connected to the hypoconid by a short, low and rounded crest. The distal marginal crest is low and indistinct, and it terminates at the base of the entoconid. There is no tuberculum sextum or other subsidiary tubercles. The distal fovea is a small triangular basin, which opens lingually at a shallow notch between the entoconid and the distal marginal crest. The size ratio of the lower molars (mean areas of M1:M2:M3) is 77:100:139.
Comparisons Although this taxon is poorly represented, it can be readily distinguished from cf. Rhinocolobus sp. based on its size and on its distinctive suite of morphological features. Although there is slight overlap in size of the lower premolars in female individuals, the lower molars in Cercopithecoides sp. are absolutely smaller (the combined molar area is 27.6% smaller on average), the upper premolars are relatively broader and higher crowned with a transverse crest that does not reach the protocone, a very reduced and more lingually positioned protocone, and a more convex distal margin, the P3 and P4 are relatively narrower and more longitudinally aligned with the long axis of the molar row, the lower molars are less hypsodont with shallower lingual notches, and M3 has more obliquely oriented lophids, a longer distal heel and a more strongly buccally off-set hypoconulid. The mandibular corpus is more robust (breadth-height index of corpus at M2 is 50.8 compared with 42.4 in cf. Rhinocolobus sp.) and it shallows slightly posteriorly rather than deepens. As noted above, the very reduced P3 protocone, the marked wear differential between the lingual and buccal cusps on the lower molars, and the specialized morphology of the mandibular corpus allow the Laetoli material to be assigned to Cercopithecoides. However, the taxon can be distinguished from all of the currently recognized species of the genus by its unique combination of features. It differs from C. williamsi in being smaller in overall dental size (see Table 6.8), having a relatively larger M3, relatively broader lower molars, and lacking prominentia laterales on the mandibular corpus. Cercopithecoides kimeui is much larger in dental size, with much broader lower molars, less marked size differential between M1 and M2, relatively much smaller M3 (the ratio of M1:M2:M3 area is 83:100:114 in C. kimeui compared with 77:100:139 in Cercopithecoides sp.), and relatively lower and more inflated cusps. The Laetoli species is similar in overall dental size to C. meavae (Table 6.8), but differs in having a narrower P3 with a shorter mesiobuccal honing face, broader and relatively larger P4, less elevated lower molar cusps with shallower lingual notches and less well-developed lophids, greater size differential between M1 and M2, a narrower and slightly smaller M3 with a shorter hypoconulid heel, and a mandibular corpus that is shallower and not as robust. Cercopithecoides kerioensis is much smaller in dental size than Cercopithecoides sp. from Laetoli, and differs in having a narrower P4 with a small metaconid, relatively broader lower molars, a relatively smaller M3 (the ratio of M2:M3 area is 100:114 in C. kerioensis, compared with 100:139 in Cercopithecoides sp.) with a reduced hypoconulid lobe, and a more slender mandibular corpus. Comparisons with C. alemayehui are hampered by the fact that few overlapping anatomical elements are preserved; only the upper premolars can be compared. The P3 is significantly smaller in
122
T. Harrison
C. alemayehui than in Cercopithecoides sp. from Laetoli. As discussed below, if the isolated postcranials from Laetoli referred to Cercopithecoides sp. are appropriately attributed, then this species is more arboreal than other species of Cercopithecoides.
Postcranial Remains of Fossil Cercopithecids from Laetoli A small collection of cercopithecid postcranials has been recovered from Laetoli (Table 6.11). All of the specimens come from the Upper Laetolil Beds, with the exception of EP 1366/01, a proximal humerus from the Lower Laetolil Beds at Emboremony, and LAET 75-415, a metatarsal fragment from the Upper Ndolanya Beds at Loc. 14. Most of the finds
are isolated and fragmentary, but two specimens are associated with cranio-dental specimens (LAET 74-247 and LAET 76-3904), and can be attributed with certainty to a particular taxon. Nevertheless, many of the other specimens can be provisionally assigned taxonomically, based on their size and morphology (see Table 6.11). Delson et al. (2000) have provided reasonable body mass estimates for Pp. ado and cf. Rhinocolobus sp. from Laetoli, based on dental dimensions (mean of 21 kg [19–25 kg] for males and 12 kg [11–13 kg] for females of Pp. ado, and mean of 34 kg [28– 41 kg] for males and 17 kg [15–19 kg] for females of cf. Rhinocolobus sp.). Leakey and Delson (1987) briefly commented on many of the postcranial specimens recovered by Mary Leakey. The aim here is to provide a brief account of the anatomy, functional morphology, and taxonomic affinities of the postcranial specimens from Laetoli, including a number of new specimens recovered since 1998.
Table 6.11 List of cercopithecid postcranial specimens from Laetoli Specimena Loc.b Horizonc Element and commentsd LAET 74-247 LAET 74-327
3 7
Lt proximal femur. Associated with cranio-dental remains of cf. Rhinocolobus sp. Rt distal femur. According to Leakey catalogue, possibly same individual as 74-322. Parapapio ado LAET 75-415 14 U. Ndolanya Rt proximal metatarsal IV LAET 75-672 1 Lt cuboid. Parapapio ado LAET 75-744 1 Lt calcaneum missing the tuber calcis. Probably cf. Rhinocolobus sp. LAET 75-1177 6 Rt talus, missing the head and neck. cf. Rhinocolobus sp. LAET 75-1817 10W Lt proximal femur. Cercopithecoides sp. LAET 75-2283 10E Lt calcaneus. Distal portion only. Parapapio ado LAET 75-3611 22 Associated rt proximal metatarsal IV and V. Parapapio ado LAET 76-3870 16 Lt talus. cf. Rhinocolobus sp. LAET 76-3904 2 Associated lt cuboid, lt capitate, lt metacarpal I, lt metacarpal V, proximal phalanx (probably lt manual phalanx IV). Associated with P4 of Parapapio ado LAET 78-4907 22S Rt proximal radius LAET 78-4911 22 6.1 m below Tuff 7 Proximal manual phalanx (probably lt ray IV). Parapapio ado LAET 78-4925 5 4.6 m above Tuff 3 Lt distal humerus. Papionin gen. et sp. indet. EP 399/98 10E Tuffs 5-7 Rt proximal humerus. Parapapio ado EP 1001/98 9S Below Tuff 2 Rt talus. Parapapio ado EP 067/99 10E Tuffs 5-7 Rt proximal humerus. Cercopithecoides sp. EP 1052/00 10 Below Tuff 2 Distal end of proximal pedal phalanx. Probably from ray V EP 210/01 3 Tuffs 7-8 Lt scaphoid. Probably cf. Rhinocolobus sp. EP 1366/01 Emb. 1 L. Laetolil Rt. proximal humerus. Cercopithecoides sp. EP 896/03 10E Tuffs 5-7 Rt proximal metatarsal IV. Cercopithecoides sp. EP 774/03 9 Tuffs 5-7 Distal end of middle pedal phalanx EP 142/04 22 Tuffs 5-7 Rt proximal radius. Parapapio ado EP 902/05 10E Tuffs 5-7 Terminal phalanx. Parapapio ado EP 963/05 2 Tuffs 5-7 Distal end of proximal phalanx. Probably lateral manual digit a Specimen prefixes: LAET, Kenya National Museum, Nairobi (on loan from Tanzanian National Museum), 1974–1979 Mary Leakey collections; EP, Eyasi Plateau Expedition, National Museum of Tanzania, Dar es Salaam, 1998–2005 Terry Harrison collections b Localities: Emb. 1, Emboremony 1. Numbers refer to the collecting localities designated by Leakey (1987a) c Horizon: The stratigraphic provenience of the Kohl-Larsen collections and most of the specimens collected by Leakey are from unknown horizons within the Upper Laetolil Beds, except where indicated. The Harrison collections are mostly surface finds, and the stratigraphic provenience is recorded as a fossiliferous section between two marker tuffs within the Upper Laetolil Beds (unless more precise provenience is known for in situ specimens). All specimens are from the Upper Laetolil Beds, except those listed as U. Ndolanya from the Upper Ndolanya Beds and L. Laetolil from the Lower Laetolil Beds d Element and comments: lt, left; rt, right. Provisional taxonomic attributions are based on size and morphological comparisons
6 Laetoli Cercopithecids
Postcranials Attributed to Parapapio
123
LAET 76-3904 consists of several postcranial elements – left cuboid, left capitate, left metacarpal I, left metacarpal V, and proximal phalanx (probably a left manual phalanx from ray IV) – associated with a P4 belonging to Pp. ado (Fig. 6.18a, Table 6.11). The postcranial elements are much smaller than
those of male Papio anubis and slightly larger than male Lophocebus aterrimus, and are therefore consistent in size with female Pp. ado. The cuboid in LAET 76-3904 is relatively long proximodistally, with narrow proximal and distal facets. It is quite similar in proportions to the cuboid of Lophocebus and Macaca, and contrasts with the relatively short cuboid seen
Fig. 6.18 Parapapio. Postcranial remains. (a) LAET 76-3904. Associated bones of left manus and left pes. Top row, left to right, metacarpal I, metacarpal V, capitate. Bottom row, left to right, metacarpal shaft fragment,
proximal phalanx (probably ray IV), cuboid. (b) LAET 78-4911, proximal manual phalanx (probably left ray IV). Left, dorsal view. Right, ventral view. (c) EP 1001/98, right talus. Left, dorsal view. Right, ventral view
124
in Papio, Nasalis and Colobus. The relatively short cuboid in extant colobines is associated with proximodistal abbreviation of the tarsus in general (Strasser 1988). In dorsal view, the bone is subrectangular, narrowing slightly proximally, with a shallow waisting of the bone towards the proximal end. Again, this matches well with the pattern seen in Lophocebus, and it can be distinguished from those of Papio and Colobus, which taper more strongly proximally, and exhibit a more marked degree of waisting. In dorsal view, the facets for metatarsals IV and V are offset at a slight angle of about 20°. In proximal view, the dorsal surface is strongly convex and the proximal articular facet is U-shaped and quite narrow, similar to the pattern seen in Lophocebus. Papio and Colobus by contrast have wider proximal articular facets with a low arc of curvature. This indicates that the foot of Pp. ado had a relatively wide range of rotation at the calcaneocuboid joint, suggestive of a relatively mobile foot in pronation-supination. The calcaneal facet protrudes slightly proximally, as in Lophocebus and Colobus, but is less protuberant than in Papio. The plantar tubercle on the ventral surface is relatively well-developed as in Papio. In distal view, the two facets for metatarsals IV and V are similar in breadth, but the latter is slightly dorsoventrally shorter due to the tapering of the bone laterally. This is similar to the configuration seen in extant cercopithecids, except that the fossil tends to have a more squared-off rather than convex lateral margin. Medially, there are confluent articular facets for the navicular and the lateral cuneiform. Along the disto-dorsal margin of the medial aspect there is a second small D-shaped facet for the lateral cuneiform. The proximal facet for the lateral cuneiform is slightly larger than the distal facet, as is typical of cercopithecines (in colobines, the proximal facet is smaller or absent) (Strasser and Delson 1987; Strasser 1988). The configuration of the facets on the medial side of the cuboid is quite similar to that in Lophocebus. Laterally, there is a shallow groove for the peroneus longus tendon, as in Lophocebus and Colobus, which is not as deep as in Papio. There is a large reniform facet on the proximal aspect of the lateral face for contact with a sesamoid (os peroneum) in the tendon of peroneus longus (Le Minor 1987; Strasser 1988). Overall, the cuboid is very similar in morphology to that of Lophocebus; the main difference being the somewhat better developed plantar tubercle. An isolated left cuboid (LAET 75-672), similar in size and morphology to the specimen described above, can also be attributed to Pp. ado. It differs, however, in being slightly shorter, having a slightly wider distal articular surface, and a more concave dorsal surface. These differences prompted Leakey and Delson (1987) to refer this specimen to the small indeterminate colobine (here recognized as Cercopithecoides sp.), but the two fossil cuboids are structurally and functionally so similar that there can be little doubt that they belong to the same taxon. Besides, LAET 75-672 has a relatively
T. Harrison
large proximal articular facet for the lateral cuneiform, typical of cercopithecines (Strasser and Delson 1987). LAET 75-2283 is a fragmentary left calcaneus, lacking much of the tuber calcis. The bone is quite heavily weathered and abraded, and there appears to be a carnivore bite mark (2 mm in diameter) on the medial side of the heel process. The calcaneus is larger than that of male Colobus guereza, and similar in size to male Lophocebus albigena, being consistent in size with female individuals of Pp. ado. Distally, the cuboid articular facet is lunate in shape, and markedly concave, with a deep ligamentous pit as in Colobus and Lophocebus. In Papio, by contrast, the facet is relatively shallower. The cuboid facet is tilted obliquely (28°) to the mediolateral plane of the sustentaculum tali, as in Papio (30°) and Lophocebus (33°), whereas in Colobus it is more horizontally aligned (10°). The distal segment of the calcaneus, anterior to the posterior talar facet, is moderately long as in extant papionins. The anterior and middle talar articular surfaces consists of two elliptical facets separated by a distance of only 1.5 mm, as in Papio and Lophocebus, whereas colobines have facets that are more widely separated. The middle talar facet forms an elongated ellipse as in papionins, rather than a small circular facet typical of colobines. There is a prominent roughened tubercle on the superior surface of the distal segment, just proximal to the cuboid facet, for the attachment of a well-developed talo-calcaneal ligament. The sustentaculum tali forms a small triangular platform. On its inferior surface there is a broad and shallow groove for the flexor hallucis longus. The posterior talar facet is short and broad, and quite elevated. The lateral margin of the calcaneus bows outwards, and is roughened for the peroneal tubercle. Inferiorly, there is a large anterior tubercle for attachment of the plantar ligament. In general, the morphology in LAET 75-2283 is most similar to papionins among extant cercopithecids in having a relatively long anterior segment, closely associated anterior talar facets, an elongated middle talar facet, an obliquely oriented cuboid facet, and a relatively large anterior tubercle on the plantar surface. It differs from terrestrial papionins, however, in having a shallower cuboid articular facet. Overall, the fossil calcaneus is a good match for that of Lophocebus albigena, and functionally it implies that Pp. ado was less specialized for terrestrial locomotion than Papio. EP 1001/98 consists of a right talus in which the medial aspect is weathered and incomplete (Fig. 6.18c). Based on its size and general morphological similarity to extant Papio, it is assigned to Pp. ado. The head is incomplete, but it appears to have had a relatively strongly convex articular surface for the navicular. A short articular tail extends from the head onto the lateral margin of the neck. The neck is short and quite robust, with a neck angle of 22° relative to the long axis of the tibial trochlear facet and 46° to the posterior calcaneal facet, more similar to Papio than arboreal cercopithecids.
6 Laetoli Cercopithecids
The cup-like facet for the medial malleolus of the tibia is deep and well defined, as in extant cercopithecines (Harrison 1982; Strasser 1988). Lateral to this facet, on the distal margin of the tibial trochlear facet is a small protuberance that acts as a bony “stop” for the distal tibia in full dorsiflexion of the foot (Harrison 1982; Conroy and Rose 1983). On the medial side, there is a weakly developed tubercle for the deltoid ligament. The tibial trochlear facet is quite broad, with a shallow midline groove and a moderate degree of posterior tapering. The lateral margin of the tibial facet is much more elevated than the medial margin, and more sharply keeled. The flange for the lateral malleolus of the fibula is relatively small. On the plantar surface the posterior calcaneal facet is quite narrow and is strongly arched. The anterior calcaneal facet is missing, and the middle facet is only partially preserved. The sinus tarsi is broad and relatively deep. A small lateral tubercle is present posteriorly. The posteromedial margin is damaged, so the development of the groove for the flexor tibialis is unknown (see Strasser 1988). There are well-developed pits on the plantar surface of the neck and on the lateral side of the talar body for strong talo-calcaneal and posterior talofibular ligaments respectively. EP 1001/98 can be distinguished from the fossil tali attributed to cf. Rhinocolobus sp. (LAET 75-1177 and LAET 76-3870) in a number of important features. These include: lateral margin of tibial trochlear facet relatively more elevated and sharply keeled; cup for medial malleolus of tibia deeper and more extensive medially; a relatively small lateral flange for the medial malleolus (larger in Papio); neck relatively shorter and less medially angled (similar angulation in Papio, but the neck is longer); the talar body narrows more strongly posteriorly (even more pronounced in Papio); the sinus tarsi is broad and deep; and the deltoid tuberosity is less prominent. Among extant cercopithecids, the talus is most similar to Papio, but it does differ in having a shallower groove on the tibial trochlear facet, presence of a lateral tail on the articular surface of the head, a more strongly arched posterior calcaneal facet, a smaller flange for the medial malleolus, and a shorter neck. The elevated lateral margin and the degree of wedging of the tibial facet, the deep cup-like depression for the medial malleolar facet, and the short lateral malleolar flange all help to enhance stability of the talocrural joint during rapid parasagittal excursions. The angulation of the talar neck may relate functionally to a more adducted neutral position for the hallux. Overall, the morphology of the talus is consistent with Pp. ado being a semiterrestrial or terrestrial cercopithecid, but somewhat less specialized in this direction than Papio. LAET 75-3611 consists of associated right proximal metatarsals IV and V. The former is more complete, preserving more than half of the shaft (total length of fragment is 42.8 mm). Metatarsal V preserves only the base of the shaft (total length of fragment is 17.4 mm). Both metatarsals are
125
slightly weathered and abraded. They correspond in size to male Papio anubis, so they could be attributed to either Pp. ado or cf. Rhinocolobus sp. based on size alone. However, morphologically they are most similar to extant papionins, which makes attribution to Pp. ado more likely. In metatarsal IV, the proximal articular surface for the cuboid is a dorsoventrally tall and rectangular facet. It has an unusual tubercle on the dorsolateral margin that is not seen in extant cercopithecids. Laterally there is a single lunateshaped facet for metatarsal V. Medially there is a pair of ovoid facets for metatarsal III, of which the superior facet is about twice the size of the inferior facet. This resembles the pattern seen in extant papionins, whereas in Colobus the two facets are united. Inferiorly there is a broad rectangular facet for articulation with a sesamoid. The interosseus pits on the medial and lateral side are relatively shallow. Based on the preserved portion of the shaft, the metatarsal appears to have been relatively robust, as in Papio and Lophocebus, compared with the relatively slender shaft in Colobus. Metatarsal V in LAET 75-3611 bears a triangular facet for the cuboid that occupies the supero-medial corner of the proximal face, similar to that in Lophocebus. Laterally there is a well-developed and rugose tuberosity, which matches that in Papio. The tuberosity is smaller in Colobus and Lophocebus. Inferiorly and laterally there is a secondary tubercle with a smooth face for articulation with a sesamoid. Medially there is a large bilobed facet for metatarsal IV, which resembles that of Papio and Lophocebus. In Colobus the facet has a much smaller inferior lobe. EP 399/98 consists of a slightly weathered right proximal humerus (Fig. 6.19 a–d). In terms of overall size, it falls in the lower end of the range for Papio anubis, and is consistent in size with Pp. ado. The head is globular, and implies relatively wide ranges of excursion at the shoulder joint in the parasagittal plane as well as in abduction-adduction. The mediolateral breadth of the head (25.0 mm) is subequal to its antero-posterior length (25.1 mm). In posterior view, the head of the humerus is egg-shaped, with a relatively broad and convex articular surface proximally in the mediolateral plane, but one that narrows and becomes less convex distally. In lateral view, the articular surface of the head has a relatively strong antero-posterior curvature. Proximally, the head extends very slightly above the level of the greater and lesser tuberosities, and the distance between the two tuberosities is quite wide (14.1 mm), which is most similar to the pattern seen in arboreal cercopithecids (Jolly 1967; Harrison 1989; Rose 1989; Larson 1993). The lesser tuberosity forms a rectangular, slightly rugose protuberance. The greater tuberosity is almost circular in outline and more rugose. The greater and lesser tuberosities are subequal in elevation, but the former is much larger (the maximum diameter of the lesser tuberosity is only 67% of the diameter of the greater tuberosity). The bicipital groove is shallow and broad compared with
126
Fig. 6.19 Parapapio. Forelimb postcranial remains. (a–d) EP 399/98, right proximal humerus. (a) posterior view, (b) medial view, (c) anterior view, (d) proximal view. (e–f) EP 142/04, right proximal radius. (e) posterior view, (f) anterior view
Papio, and it is located relatively far laterally. It is bordered medially by a sharp keel that runs the length of the preserved portion of the shaft. The configuration of the articular surface of the head and the tuberosities indicate that Pp. ado was less specialized for terrestrial locomotion than extant Papio, being most consistent with semi-terrestrial or arboreal cercopithecids. EP 142/04 represents the proximal end of a right radius (Fig. 6.19e–f). The specimen is slightly smaller than those of male individuals of Lophocebus albigena, but is quite similar morphologically. It is much smaller and differs in a number of important respects from LAET 78-4907 assigned to cf. Rhinocolobus sp. Attribution to a female individual of Pp. ado is reasonable given the size and morphology. The head is ovoid in shape, with a maximum diameter of 13.7 mm. and a perpendicular breadth of 12.5 mm. The breadth-length index of the radial head is 91.2, similar to that in extant cercopithecines and less elliptical than in colobines (Harrison 1989; Ciochon 1993). The articular surface of the head is tilted
T. Harrison
obliquely towards the antero-lateral side at an angle of 10° to a plane perpendicular to the long axis of the shaft. It has a deep, circular depression placed centrally on the proximal articular surface for articulation with the capitulum of the humerus. The rim surrounding the depression is broadest and most rounded antero-laterally. The posteromedial rim is narrower, and raised to form a sharply elevated lip. The tilt of the head and the raised rim provide enhanced stability of the elbow joint in semi-flexed and pronated positions (Harrison 1989; Rose 1993). The collar of the radial head forms a rim of uneven thickness, being thickest antero-medially (4.8 mm) and shallowest laterally (1.8 mm) (see Rose 1993). The neck is relatively slender and elliptical in cross-section. It has a maximum diameter of 9.4 mm and a perpendicular breadth of 7.8 mm. The length of the neck, from the inferior rim of the head to the superior margin of the bicipital tuberosity, is relatively short (6.5 mm). The index of radial neck length (length of the neck x100/maximum breadth of the head) is 47.4, which falls within the ranges for Macaca, Papio, and Mandrillus, and is less than in Lophocebus, colobines and cercopithecins (which have relatively longer radial necks) (Harrison 1989). The bicipital tuberosity is represented by an ovoid roughened protuberance, with a slit-like central depression. The shaft below the tuberosity is subcircular in cross-section. The long-axis of the shaft is directed at an angle of 16° to the long-axis of the neck, implying a relatively bowed radial shaft. Curvature of the radius shaft tends to be pronounced in arboreal cercopithecids, associated with relatively well-developed forearm rotators (Rose 1993; Ciochon 1993). Few features serve to distinguish extant cercopithecids in the proximal radius, so it has proved difficult in the past to deduce functional differences (Jolly 1967; Birchette 1982). Nevertheless, the proportions of the head, the length of the radial neck, and the bowing of the shaft are of some significance in this regard (Harrison 1989). EP 142/04 is most similar to Lophocebus, except that the neck is relatively longer in the latter. It differs from that of Colobus in having a more robust shaft, which is less strongly laterally bowed, and in having a shorter radial neck. Compared to that of Papio, the capitular depression on the head is deeper in EP 142/04, the neck is slightly longer, the bicipital tuberosity is less protuberant, and the shaft appears to be more strongly bowed laterally. The left capitate of LAET 76-3904 is generally similar in morphology to that of Lophocebus. It differs primarily in being relatively proximo-distally longer, having a less globular scaphoid facet and less concave facet for metacarpal III. On the medial side of the distal end, there is a single continuous B-shaped facet for metacarpal II as in Lophocebus, whereas in Papio and Colobus it is subdivided into separate superior and inferior facets. The facet for metacarpal III on the distal aspect of the bone is strongly concave, most similar
6 Laetoli Cercopithecids
to Lophocebus, but relatively more pronounced. Ventral tapering of the metacarpal III facet is marked, being most similar in this respect to Papio. The palmar tubercle is weakly developed, as in Lophocebus. The articular facet for the hamate is only slightly convex. Metacarpal I in LAET 76-3904 shows that the first manual ray in Pp. ado was relatively long and well-developed. The metacarpal is 23.5 mm long. The metacarpal length:capitate proximodistal length ratio in LAET 76-3904 is 1.87, compared with 1.69, 1.61 and 1.59 in Papio, Lophocebus and Colobus respectively. The presence of a relatively long first metacarpal adds support to the attribution of these associated postcranials to Pp. ado rather to one of the colobines, which presumably had reduced thumbs as in extant colobines. The distal articular surface for the proximal phalanx is low and broad, and narrows slightly dorsally. It is dorsoventrally shallower than in Lophocebus and Papio. The shaft is relatively stout compared to extant cercopithecids. The proximal articular surface is relatively broad with a strong mediolateral convexity, most similar to Lophocebus, suggesting a wide range of abduction of the thumb. It lacks the pronounced palmar beak of Papio. Overall, the impression is that Pp. ado had a well-developed and mobile thumb that was capable of wide ranges of motion in flexion-extension, and especially in abduction-adduction. Only the proximal end of metacarpal V in LAET 76-3904 is preserved. The proximal facet is more similar to Colobus than Papio and Lophocebus in being relatively broad. The dorsoventral height and mediolateral breadth of the facet are 8.7 mm and 6.6 mm respectively. It differs from Papio in having a greater mediolateral convexity of the hamate facet and a relatively longer facet for articulation with metacarpal IV. The proximal phalanx in LAET 76-3904 is probably from ray IV of the left manus. It is short and stout, with an index of phalangeal robusticity (mid-shaft breadth × 100/maximum length of phalanx) of 25.1. This falls outside the range of extant cercopithecids, with the exception of Theropithecus gelada (Harrison 1989). As noted by Jolly (1967, 1972) and Harrison (1989), relative phalangeal length is a good indicator of substrate utilization in cercopithecids, with short stout phalanges, such as those in Pp. ado, implying a high degree of terrestriality in their locomotor behavior. An index relating the length of proximal phalanx IV against the length of the first metacarpal discriminates terrestrial forms with long thumbs and short lateral digits, such as Theropithecus gelada (62.0), Erythrocebus patas (81.9), and Papio anubis (86.6), from arboreal and semi-terrestrial cercopithecids with short thumbs and long lateral digits, such as Chlorocebus aethiops (120.2), Lophocebus spp. (121.0), Cercopithecus mitis (123.1) and Colobus guereza (136.3). The index in the fossil is 83.9, most similar to terrestrial cercopithecids. The distal articulation is dorsoventrally low and narrow, being mediolaterally narrower than the midshaft diameter. It does not
127
narrow dorsally, being most similar to Papio in this respect, implying joint stability in both flexion and extension. The midline groove of the trochlea of the distal articulation is relatively shallow, as in Lophocebus, and much shallower than in Colobus and Papio. The shaft is broad and quite flattened dorsoventrally, with a slight degree of curvature comparable to that in Papio. The flanges on the plantar surface for the flexor sheaths are strongly developed and located far distally, as in Papio. The proximal end is relatively broad, again as in Papio. Overall, the phalanx is most similar to those of terrestrial cercopithecids. LAET 78-4911 is an isolated proximal phalanx, again probably from the ray IV of the left manus (Fig. 6.18b). It is morphologically similar to the previous specimen, but is somewhat larger, being comparable in size to the corresponding phalanx in male Papio anubis, and less robust. It is identifiable as a manual phalanx by its relatively narrow distal articulation with a shallow trochlear groove. The index of phalangeal robusticity is 19.7, which falls in the upper extreme of the range for arboreal cercopithecids, but most closely corresponds to terrestrial monkeys (Harrison 1989). The distal articular surface is relatively narrow, with a shallow trochlear groove, especially on the palmar aspect. The slight degree of curvature of the shaft matches that of Papio. The palmar surface of the shaft has long and deep grooves bordered by narrow crests extending along the medial and lateral sides for attachment of the flexor sheaths. These flanges are similar in location to those in Papio, but are longer and more strongly developed in the fossil. The proximal articular surface is similar to that in Papio, except that it is slightly narrower with a greater mediolateral concavity. Overall, the stout shaft and strong flexor sheaths are features that the fossil shares with the phalanges of Papio, while the shallow trochlear groove of the distal articulation and the more concave proximal articular facet are similar to arboreal cercopithecids, such as Colobus and Lophocebus. On balance, the morphology implies that Pp. ado was a semi-terrestrial form, less committed to terrestriality than Papio. Trails of cercopithecid footprints are known from three sites (C, D, and F) at Laetoli Locs. 7, 11 and 10E respectively (Leakey 1987b). A total of 21 prints from the foot were complete enough to measure, and these ranged in length from 130–185 mm (from toe tip to center of heel; the total foot length would have been somewhat greater). This falls just below or in the lower end of the range of total foot length of extant adult male and female Papio anubis (152–229 mm for total hindfoot length; Berger 1972). Several prints also show clear impressions of the thumb. They closely resemble those of prints made by modern-day yellow baboons (Papio cynocephalus) (Bird 1987). Although there is some variation in overall size (exaggerated by the preservation), the prints are consistent with belonging to a single species. Based on size, the presence of a sizeable thumb, and their overall similarity
128
to those of extant baboons, the prints were most likely made by Pp. ado. They confirm that at least one species of cercopithecid at Laetoli, presumably Pp. ado, spent time moving quadrupedally on the ground. In summary, the postcranial specimens attributed to Pp. ado, when viewed collectively, exhibit a mosaic of features that are found in extant arboreal and semi-terrestrial papionins. It clearly was not as specialized for terrestriality as Papio or Theropithecus. Some aspects of the foot, especially the morphology of the cuboid and calcaneus, most closely resemble Lophocebus. The functional anatomy of the proximal humerus and proximal radius indicates specialization of the forelimb for semi-terrestrial and arboreal behaviors. In contrast, the talus and especially the robusticity of the phalanges indicate a greater degree of specialization for terrestrial quadrupedalism. The footprints at Laetoli provide additional support for this type of behavior. Overall, the balance of evidence suggests that Pp. ado was a relatively slender and agile semi-terrestrial monkey (generally similar, in terms of positional behavior, to Cercocebus and some species of Macaca), which was adept in the trees (where it probably spent the majority of its time), but frequently traveled on the ground to forage and to move between woodland and forest patches. Given its size and the short digits, Pp. ado would probably have preferred walking and running on large diameter supports in arboreal settings, but it was certainly capable of climbing into the upper part of the canopy to forage and to sleep at night. Postcranial material, presumably of Parapapio, has been recovered from Plio-Pleistocene localities in South Africa, but they are all unassociated and have not yet been attributed taxonomically. Based on ecomorphological comparisons, Elton (2001) suggested that the different species of Parapapio in South Africa might have had different ecological preferences, with Pp. jonesi being more terrestrial than the larger P. broomi. The postcranial morphology (i.e., distal humerus and proximal femur) in Pp. cf. jonesi from Hadar is most comparable to that of extant arboreal macaques and mangabeys (Frost and Delson 2002), and was probably similar in its inferred locomotor behavior to that presented here for Pp. ado. Additional work is needed to document the diversity of positional behavior in Parapapio spp., but at least two species from the Pliocene of East Africa, Pp. ado (Laetoli) and Pp. cf. jonesi (Hadar), were apparently semiterrestrial papionins, more arboreally adapted than Papio, Theropithecus and Mandrillus, but still highly proficient as terrestrial quadrupeds.
Postcranials Attributed to the Large Papionin LAET 78-4925 is a left distal humerus (Fig. 6.20). It is somewhat larger than Papio anubis specimens (with average linear dimensions that exceed that of a large male anubis
T. Harrison
Fig. 6.20 Papionin gen. et sp. indet. LAET 78-4925. Left distal humerus. (a) anterior view, (b) posterior view
baboons by more than 12%). The size and morphology of the specimen indicate that it belonged to a large species of papionin, as initially suggested by Leakey and Delson (1987). However, it differs from Papio in the following respects: the distal articular surface is mediolaterally broader in relation to its proximo-distal height; the medial trochlear keel is less projecting distally (the index of relative flange projection, following Frost and Delson (2002), is 59.0 in LAET 78-4925, which is intermediate between extant colobines and terrestrial cercopithecines), and its medial margin is aligned more obliquely to the long-axis of the shaft; the lateral trochlear keel is better defined (although still relatively low), and positioned more medially; the capitulum is similar in degree of convexity, but the articular surface is more expanded distally, allowing greater stability in extended positions of the elbow; the medial epicondyle is slightly shorter and more posteriorly directed (directed posteriorly at an angle of 65° to the transverse axis of the distal articulation; this falls outside the values of extant arboreal cercopithecids, just within the range of semi-terrestrial monkeys, but close to the mean values for terrestrial cercopithecines, such as Papio and Theropithecus [Harrison 1989; Frost and Delson 2002]); the lateral epicondyle is better developed; the trochlear keel projects more distally in relation to the medial epicondyle; the olecranon fossa is proximodistally lower and it lacks the perforation commonly seen in Papio, thereby limiting full extension of the elbow joint; the fossa is not excavated under the lateral olecranon margin; the medial border of the olecranon is wider than the lateral border, whereas in Papio it is the opposite; the shaft is more stoutly constructed; the lateral supracondylar ridge, which runs along the length of the shaft fragment,
6 Laetoli Cercopithecids
129
is more weakly developed; the coronoid fossa is slightly shallower; and the radial fossa is not perforated. The size and morphology of LAET 78-4925 clearly points to it belonging to a large terrestrial cercopithecid. The key functional features that indicate terrestrial behavior are as follows: projecting and sharply developed medial trochlear keel, a narrow trochlea, a less globular capitulum, a short and posteriorly directed medial epicondyle, a short lateral epicondyle, a deep and well delimited olecranon fossa, shallow radial and coronoid fossae, and a weakly developed lateral supracondylar keel (Jolly 1967, 1972; Harrison 1989; Ciochon 1993). However, it is less specialized in this direction than extant Papio, especially in the less projecting medial trochlear keel, the greater development of the capitular surface distally, and the less specialized morphology of the olecranon fossa for stabilizing the elbow joint in maximal extension.
Postcranials Attributed to cf. Rhinocolobus sp. LAET 74-247 is a left proximal femur, lacking most of the shaft (Fig. 6.21a, b). The proximal femur is associated with cranio-dental specimens of cf. Rhinocolobus sp., and this seems to be a reasonable association based on its size and morphological characteristics. The maximum diameter of the head (24.8 mm) falls in the upper end of the range of male Papio anubis and Nasalis larvatus, and is, therefore, consistent in size with cf. Rhinocolobus sp. The superior and inferior surfaces of the head are mediolaterally convex, with a marked extension of the articular surface laterally onto the dorsal surface of the femoral neck. There is a well-developed fovea capitis, 11.0 mm by 6.2 mm in diameter, which is relatively centrically placed. In anterior view, the head was positioned almost at the level of the apex of the greater trochanter, being only 1.9 mm lower. The height of the greater trochanter above the femur neck in relation to the diameter of the head yields an index of 27.8, most similar to modern-day arboreal colobines (Harrison 1982; Frost and Delson 2002; Frost 2007). The degree of projection of the greater trochanter influences the mechanical advantage of the gluteus medius and piriformis muscles acting about the hip joint (Smith and Savage 1956). A relatively low projection of the greater trochanter, as seen in LAET 74-247 and extant colobines, permits rapid extension of the hindlimb, important in arboreal leaping (Walker 1974). The neck is moderately short, as in extant colobines, with the marginal lip of the head being separated from the superior margin of the greater trochanter by 11.7 mm and from the inferior margin of the lesser trochanter by 24.1 mm. The neck is quite robust, with a minimum antero-posterior breadth of 13.9 mm. A short and stout femoral neck is an adaptation
Fig. 6.21 cf. Rhinocolobus sp. Postcranial remains. (a, b) LAET 74-247, left proximal femur. (a) posterior view, (b) anterior view. (c, d) LAET 76-3870, left talus. (c) dorsal view, (d) ventral view. (e–h) EP 210/01, left scaphoid. (e) proximal view, (f) distal view, (g) medial view, (h) lateral view
to resist greater bending stresses during arboreal leaping and climbing (Fleagle 1977; Fleagle and Meldrum 1988; Harrison and Harris 1996). The neck angle is difficult to measure precisely because much of the shaft is lacking, but it can be estimated to have been 123° relative to the long axis of the shaft, being most similar to extant colobines and arboreal cercopithecins (Harrison and Harris 1996; Frost and Delson 2002; Frost 2007). This feature is presumably functionally related to increased use of the hindlimb in abducted positions during arboreal climbing (Harrison 1982; Harrison and Harris 1996). The specimen differs from the proximal femur of Papio in having a more globular femoral head, a relatively lower greater trochanter in relation to the head, a slightly shorter neck with a greater neck angle, and a less prominent lesser trochanter. These features indicate that the femur belonged to a more arboreally adapted cercopithecid. Morphologically, it is quite similar to Colobus and Piliocolobus, but differs in having a slightly lower greater trochanter and a less globular
130
femoral head. LAET 75-247 differs from the proximal femur of Paracolobus mutiwa (KNM-WT 16827), Cercopithecoides williamsi (KNM-ER 4420) and Cercopithecoides meavae (A.L. 2–70) in having a shorter and more robust neck, a higher neck angle, and a much less elevated greater trochanter. It is much more similar to Paracolobus chemeroni (KNM-BC 3), which is slightly larger, but corresponds in having a short neck and a low greater trochanter in relation to the position of the head. It differs, however, in having a higher neck angle. It appears to have been more specialized for arboreal behaviors than most of the other PlioPleistocene East African large colobines, with the exception of Paracolobus chemeroni (Birchette 1982). LAET 74-327 is a right distal femur lacking the medial condyle. It is of comparable size to thsssse previous specimen, being as large as an extant male Papio anubis. The general morphology is comparable to that of Colobus in being relatively antero-posteriorly short in relation to the mediolateral width, and in having a low and broad patella groove, a low rounded lateral keel to the patella groove, and a lateral condyle that narrows distally. These features all indicate that the joint was less effectively stabilized during rapid flexionextension at the knee compared to terrestrial cercopithecids. LAET 76-3870 consists of a complete, but slightly weathered and abraded, left talus (Fig. 6.21c, d). It is slightly larger than the talus of a male Papio anubis, and thus conforms to the expected size for cf. Rhinocolobus sp. LAET 75-1177 is a more fragmentary right talus, lacking the head and neck, as well as the proximal portion of the talar body. Morphologically it is very similar to LAET 76-3870, but slightly smaller (the linear dimensions are on average 10.9% smaller). The head of the talus has a strong mediolateral curvature as in extant colobines, and extends more medially than in Papio and Lophocebus. In distal view, the mediolateral longaxis of the head is relatively strongly obliquely tilted in relation to the subtalar plane (20° in LAET 76-3870) as in Colobus, and much more so than in terrestrial cercopithecids, such as Papio. The neck is moderately long and quite robust, being more similar in this respect to Papio and Lophocebus, than to Colobus, which has a relatively shorter neck. The talar neck in LAET 76-3870 is oriented mediolaterally at an angle of 34° to the long axis of the tibial articular facet and 63° to the posterior calcaneal facet. The head and neck are directed more medially than in Papio and Lophocebus, but are comparable to the orientation in Colobus. The orientation of the head and neck in LAET 76-3870 implies that the foot was well adapted for inverted positions during climbing. There is a well-developed cup-like articular facet on the medial side of the neck for articulation with the medial malleolus of the distal tibia. It is similar to that in Colobus and Lophocebus, and slightly shallower than that in Papio. The articular surface for the distal tibia has a relatively low and rounded lateral margin and a broad, shallow
T. Harrison
trochlear groove, as in arboreal cercopithecids, such as Colobus and Lophocebus. In Papio, by contrast, the lateral margin of the tibial facet is sharp and more elevated, with a deeper trochlear groove. The tibial facet tapers proximally with a low degree of wedging (index of wedging = breadth of the tibial facet proximally × 100/ breadth of the tibial facet distally, including the malleolar cup; index = 65.8 in LAET 76-3870, compared to mean values of 50.8 in Papio anubis, 55.4 in Cercopithecus aethiops, and 54.9 in Colobus guereza; Harrison 1982). Wedging of the tibial facet is highest in terrestrial cercopithecids to maintain greater stability of the talocrural joint during parasagittal excursions, especially in positions of extreme dorsiflexion. The fibular flange on the lateral side of the talus for articulation with the malleolus of the fibula is only moderately developed, being intermediate between the short flange of Colobus and Lophocebus, and the strongly developed flange in Papio. The ligamentous pits on the posterolateral and posteromedial aspects of the talar body for the posterior talofibular and posterior talotibial ligaments are well developed compared with those in Colobus, but are not as strongly pitted as in Papio. The tubercle for the deltoid ligament is prominent as in Colobus, and more strongly developed than in Lophocebus and Papio. Proximally, the groove for the tendon of the flexor hallucis longus is broad and shallow compared with that in Papio and Lophocebus, and is more similar to the configuration in Colobus. There is a well-developed medial tubercle and a low lateral tubercle. On the plantar surface, the sulcus tali is quite deep and broad. The middle calcaneal facet is long and narrow, and gently convex. The posterior calcaneal facet is relatively wide, strongly concave dorsoventrally, and in LAET 76-3870 its long-axis is oriented at an angle of 83° to the long-axis of the tibial articular surface. LAET 76-3870 and LAET 75-1177 share a number of key structural features with Colobus that distinguish it from terrestrial cercopithecids. These include: talar head with strong mediolateral convexity and marked dorso-ventral tilt; shallow articular depression for the medial malleolus of the tibia; relatively narrow flange for the lateral malleolus of the fibula; tibial articular surface that is only slightly wedged, with shallow trochlear groove, and less elevated lateral margin; shallow and broad groove for the flexor hallucis longus; well-developed deltoid tubercle; and moderately developed medial and lateral ligamentous pits. These features indicate that the foot of Rhinocolobus was similar to Colobus in being less specialized for stability during fast, parasagittal actions of the talocrural joint, and better adapted for inversion of the foot in arboreal climbing. LAET 75-744 consists of a left calcaneus lacking the proximal end of the heel process. The specimen is slightly larger in size than male Papio anubis, and is much larger in size than LAET 75-2283 assigned to Pp. ado. Based on the
6 Laetoli Cercopithecids
size of the specimen an attribution to cf. Rhinocolobus seems reasonable. The cuboid facet is quite deep, as in Colobus and Lophocebus, compared with the relatively shallow depression in Papio. Its mediolateral alignment (6°) in relation to the plane of the sustentaculum tali, is much more similar to that of Colobus (10°), than to Papio (30°) and Lophocebus (33°). The anterior segment of the calcaneus distal to the posterior facet is moderately long, as in Papio and Lophocebus, and longer than in Colobus. Disposition of the anterior talar articular surface is obscured by weathering. The middle articular facet is long and elliptical as in cercopithecines, rather than subcircular as in colobines. The peroneal tubercle on the lateral side is massively developed, most similar to that of Papio. The sustentaculum tali is mediolaterally relatively short. Inferior to the sustentaculum tali is a shallow groove for the flexor hallucis longus. Inferiorly, there is a weakly developed anterior tubercle. The heel process is short and relatively deep dorsoventrally. Superiorly, a well-developed ridge runs from the posterior margin of the posterior talar facet onto the superior aspect of the heel process. LAET 74-744 exhibits a mosaic of features found in extant cercopithecids, but several key aspects of the anatomy link it structurally and functionally to the calcanei of colobines: the low angle of mediolateral tilt of the cuboid articulation, the weakly developed anterior tubercle, and the short and deep heel process. It differs from the calcaneus of Paracolobus mutiwa (KNM-WT 16827) in being slightly smaller in overall size, and in having a slightly less concave cuboid facet, a smaller and less proximally positioned peroneal tuberosity, a narrower sustentaculum tarsi and a deeper heel process, but overall it is functionally similar. EP 210/01 is a left scaphoid, missing the tip of the mediodistal beak (Fig. 6.21e–h). It is slightly larger in overall size than male specimens of Papio anubis, so it is consistent in size with the other postrcranial elements of cf. Rhinocolobus sp. The radial facet is dorsoventrally relatively wide, and reniform in shape. The inferior margin of the facet is smoothly rounded, while the superior margin is lipped. The remainder of the proximal surface to the medial side of the radial facet is a roughened subcutaneous surface, which forms a rounded, but stout prominence medially for the attachment of the abductor pollicis brevis. Distally, there is a crescent-shaped facet on the lateral margin for articulation with the lunate. Medial and inferior to the lunate facet is a deep, tear-drop shaped concavity for the capitate. Superior to the capitate facet is a smooth articular surface for the os centrale. Medially there is a distally directed beak-like process, which, although incomplete, was clearly a well-developed process. The superior face of the beak has the remnant of an elliptical facet for articulation with the trapezoid. Compared with the similar-sized scaphoid in Papio, EP 210/01 differs in having a mediolaterally broader and more convex articular surface for the radius, a more concave facet for the capitate,
131
and a stouter proximo-medial tubercle for attachment of the abductor pollicis brevis. These features suggest that cf. Rhinocolobus sp. had a greater range of abduction-adduction of the wrist, a more mobile scaphoid-capitate joint, and an enhanced capability for abduction of the thumb. The postcranials attributed to cf. Rhinocolobus sp. are readily distinguishable from those assigned to the two species of papionins at Laetoli, and they most closely resemble those of extant colobines, as well as some of the large extinct colobines from the Mio-Pliocene of East Africa. Although the material is too scanty to reconstruct its locomotor behavior, the large colobine from Laetoli was evidently specialized for arboreal quadrupedalism, including adaptations for climbing and leaping. The postcranials are generally similar to those of Rhinocolobus turkanaensis, Paracolobus chemeroni and Paracolobus enkorikae, which were apparently predominantly arboreal quadrupeds (Birchette 1982; Leakey 1982; Ciochon 1993; Frost and Delson 2002; Hlusko 2007), and they differ from those of the more specialized terrestrial fossil colobines belonging to Cercopithecoides spp. and Pc. mutiwa (Birchette 1982; Leakey 1982; Harris et al. 1988; Frost and Delson 2002; Jablonski et al. 2008b; Harrison, unpublished). Unfortunately, there are no overlapping skeletal elements to compare with Kuseracolobus, but the postcranial morphology does indicate that, like Rhinocolobus and Paracolobus, Kuseracolobus was predominantly arboreal (Hlusko 2006).
Postcranials of Cercopithecoides sp. LAET 75-1817 is a left proximal femur (Fig. 6.22g, h). It is slightly smaller than the femur of female individuals of Colobus guereza, and, therefore, consistent in size with Cercopithecoides sp. The head is relatively globular, as in Colobus and Lophocebus, with a more extensive articular surface than Papio. The head is relatively small (diameter = 15.1 mm) in relation to the stout shaft, as in Colobus. The maximum diameter of the fovea capitis is 6.6 mm. The neck is relatively long, as in arboreal and semi-terrestrial cercopithecines. The length of the neck from the base of the head to the proximal margin of the lesser trochanter is 11.1. It is quite stout, with a minimum antero-posterior and mediolateral diameter of 8.8 and 12.5 respectively. The neck angle is 128° relative to the long-axis of the shaft, similar to that of extant colobines and arboreal cercopithecins (Frost 2007). Papio, Mandrillus and terrestrial cercopithecins differ in have a shorter neck with a lower neck angle (Harrison and Harris 1996; Frost and Delson 2002). The greater trochanter is not preserved, but it probably extended proximally to a level slightly above that of the head. There is a well-developed intertrochanteric line that reaches the lesser trochanter as in
132
Fig. 6.22 Cercopithecoides sp. (a–c) EP 067/99, right proximal humerus. (a) posterior view, (b) medial view, (c) proximal view. (d–f) EP 1366/01, right proximal humerus. (d) posterior view, (e) medial view, (f) proximal view. (g–h) LAET 75-1817, left proximal femur. (g) anterior view, (h) posterior view. Top scale bar refers to a–f; bottom scale bar refers to g–h
Papio and Cercocebus, and is more distinct than in Colobus. Unlike Colobus, there is no indication of an intertrochanteric tubercle. The lesser trochanter is prominent, providing a mechanical advantageous insertion for the iliacus. Inferior to the lesser trochanter is a well-developed crest for the vastus medialis and pectineus muscles. On the posterior surface of the shaft, just distal to the greater trochanter, there are strong muscle markings leading down to the linea aspera, presumably for attachment of the quadratis femoris and adductor brevis muscles (Howell and Straus 1933). The shape of the head and orientation and length of the neck indicate a wide range of abduction and rotation at the hip, typical of arboreal cercopithecids. This inference is further supported by the strong muscle markings for thigh abductors, adductors and lateral rotators in the fossil. Overall, the impression is one of a small arboreal cercopithecid, quite similar in its overall morphology to extant African colobines. EP 067/99 is a slightly weathered right proximal humerus (Fig. 6.22a–c). It is similar in size to male Lophocebus
T. Harrison
albigena and slightly larger than male Colobus guereza, and consistent in size either with female Pp. ado or Cercopithecoides sp. However, it is much smaller (average linear measurements are 23.7% smaller) and morphologically distinct from EP 399/98 assigned to Pp. ado, and it shares key features with extant colobines that indicate that the specimen is best attributed to Cercopithecoides sp. The head is globular, with a sub-circular shape in posterior view, and a strong degree of convexity in both the mediolateral and anteroposterior planes (see Harrison 1989). Compared with EP 399/98 of Pp. ado, the head is much more spherical, with a greater curvature of the articular surfaces. Estimated maximal ranges of motion in each plane (based on the angle of convergence of lines perpendicular to the margins of the articular surface) are 144° in the antero-posterior plane and 110° in the mediolateral plane. This compares to values in EP 399/98, assigned to Pp. ado, of 123° and 62° respectively. The articular surface of the head extends proximally well above the greater and lesser tuberosities (at least 2 mm), as seen in arboreal cercopithecids (Harrison 1989). Clearly, the humero-scapular joint was highly mobile and accommodated a wide range of flexion-extension, abduction-adduction and circumduction, typical of arboreal cercopithecids. The greater and lesser tuberosities are relatively low, elliptical protuberances. The greater tuberosity is slightly more elevated proximally than the lesser tuberosity. The bicipital groove is shallow and ill-defined. EP 1366/01 is a right proximal humerus that is rather badly weathered, but the main morphological features are still discernable (Fig. 6.22d–f). This is the only cercopithecid specimen known from the Lower Laetolil Beds (from Emboremony 1). It is similar in size and morphology to EP 067/99, and it likely belongs to the same species. The head is globular with strong convexity of the articular surface in the mediolateral and proximodistal planes. Estimated maximal ranges of motion in each plane (based on the angle of convergence of lines perpendicular to the margins of the articular surface) are 144° in the antero-posterior plane and 95° in the mediolateral plane, similar to those of EP 067/99. The head projects proximally 6 mm above the level of the greater tuberosity, so is even more elevated than in EP 067/99. The bicipital groove is broad and shallow. It differs slightly from EP 067/99 in the shape of the head, in that it is relatively longer in the proximodistal diameter than in the mediolateral breadth, and the head narrows slightly posteriorly, rather than being sub-circular. EP 896/03 consists of the proximal end and much of the shaft of a right metatarsal IV. The proximal articular surfaces are slightly abraded. The fossil is identical in size and similar in morphology to the corresponding element in male individuals of Colobus guereza. Based on its size it can be attributed with confidence to Cercopithecoides sp. The cuboid facet is not fully preserved, but it was evidently a rectangular
6 Laetoli Cercopithecids
facet that was dorsoventrally taller than wide. It is generally quite strongly convex in the dorsoventral plane, and the facet is tilted at an angle of 18° to the vertical plane relative to the long-axis of the shaft, so that it faces slightly dorsally. The facet also has a 10° mediolateral obliquity, with the medial margin being more distally placed than the lateral margin. A long and narrow crescent-shaped facet occurs on the lateral side of the proximal end for articulation with metatarsal V. It is dorsoventrally concave. Superiorly, the facet has a distinct beak, which allows greater stability at the metatarsal IV-V joint. Medially, the proximal end preserves the remnants of two distinct facets for metatarsal III. The superior facet is preserved almost intact. The inferior facet is abraded, leaving only a roughened ridge. The shaft is slender with a relatively strong dorsoventral curvature and a slight degree of curvature to the lateral side. The postcranial specimens attributed to the smaller species of colobine from Laetoli are morphologically very similar to those of extant Colobus, and clearly indicate that this taxon was fully arboreal in its locomotor behavior. If this is the case, Cercopithecoides sp. from Laetoli contrasts with C. williamsi, C. kimeui and C. meavae, which all appear to have postcrania that indicate a relatively high degree of terrestrial specialization (Birchette 1982; Leakey 1982; Elton 2001; Frost and Delson 2002; Jablonski et al. 2008b). It is conceivable that the Laetoli species represents the primitive sister taxon of these three species (consistent with its greater antiquity), and that it retains the primitive postcranial morphology and arboreal habitus that was subsequently modified to allow increased terrestriality in the last common ancestor of later species of Cercopithecoides. This hypothesis could be tested with the recovery of postcranial material of Cercopithecoides kerioensis from Lothagam, which may represent the earliest and most primitive known member of this lineage (Leakey et al. 2003). Alternatively, if the postcranial distinctions are confirmed by additional associated discoveries, they might provide sufficient evidence to recognize the Laetoli taxon as a separate genus.
Unassigned Cercopithecid Postcranials LAET 78-4907 is a right proximal radius. It is comparable in size to male Papio anubis, so it could be assigned to either Pp. ado or cf. Rhinocolobus based on size alone. The maximum length of the fragment is 45.3 mm. The head is elliptical in shape, with a shallow central depression, and a raised posterior lip. Although the head is weathered and abraded, it is possible to estimate its maximum diameter and perpendicular breadth at 21.0 mm and 17.3 mm respectively. The breadth-length of the radial head is 82.4, which indicates a relatively elliptical shape, most similar to extant colobines
133
(Harrison 1989). The antero-posterior tilt of the head is quite marked (9°). The tilt of the head and the lipping provide for greater stability of the humero-radial joint in semi-flexed and pronated postures (Harrison 1989; Rose 1993). The neck is slender and appears to be relatively short. The bicipital tuberosity is well developed and rugose as in Papio and Lophocebus, and better developed than in Colobus. There is a strong ridge and depression on the antero-lateral side of the shaft for the attachment of the supinator. LAET 75-415 consists of the proximal end and shaft of a right metatarsal IV. This is the only postcranial element of a cercopithecid known from the Upper Ndolanya Beds (Loc. 14). It is larger and more robust than the corresponding bone in Papio. Comparisons with LAET 75-3611, assigned to Pp. ado, show that LAET 75-415 is much larger (linear dimensions are almost 27% larger) and more robust. The specimen could belong to cf. Rhinocolobus sp., which has been provisionally identified from the Upper Ndolanya Beds, but given that we know so little about the cercopithecid community from this stratigraphic unit, it is probably best to leave the specimen unassigned. The proximal end is eroded, so the details of the articular surfaces for the cuboid and metatarsals cannot be determined. The cuboid facet is subrectangular in shape. The elliptical superior facet for metatarsal III is separated from the cuboid facet by a distance of 1.6 mm. The inferior facet for metatarsal III is not preserved. Inferiorly there is a small rectangular facet for a sesamoid. The facet for metatarsal V on the lateral side is only preserved superiorly. Judging from its eroded base, it was a bilobed facet. The shaft appears to be very stout, being more robust even than in Papio. EP 1052/00 is the distal end and portion of the shaft of a pedal proximal phalanx, probably from ray V. The distal end is relatively broad, with a trochleiform articular surface that narrows dorsally. The midline groove is quite deep. The shaft appears to have been relatively slender, with minimal dorsoventral curvature, and weak development of the flanges for the flexor sheaths. It is most similar to the morphology seen in Lophocebus albigena, except that the shaft is more slender and the flexor sheaths are less pronounced. EP 963/05 is the distal end of a manual proximal phalanx from a lateral ray. The distal articular surface is relative broad and narrows dorsally, with a moderately well-developed midline groove. Medially and laterally, there are shallow depressions for the collateral ligaments. The shaft is dorsoventrally flattened, with sharp keels medially and laterally for the attachment of the flexor sheaths. EP 774/03 is the distal end of a middle phalanx from the pes. The distal end is relatively broad, with a wide V-shaped trochlear groove separating the articular condyles. The shaft is quite slender, dorsoventrally compressed, and elliptical in section. It is very slightly dorsoventrally curved. It differs from Papio in having a dorsoventrally shallower distal articulation
134
and a more slender shaft, and differs from Colobus in having a less curved shaft. Overall, like the previous three specimens, it is metrically and morphologically similar to the corresponding phalanges of Lophocebus albigena. It is likely that all of these phalanges belong to Pp. ado based on size alone, but without associated comparative material an attribution to cf. Rhinocolobus sp. cannot be entirely discounted. EP 902/05 is a complete, slightly weathered terminal phalanx. It is from a lateral digit, but it is not possible to discern whether it is from the pes or manus. It is intermediate in size between those of male Colobus guereza and Papio anubis. It is a relatively quite long and slender phalanx. The proximal end is mediolaterally narrow. The articular surface for the middle phalanx is sellar, being dorsoventrally concave and mediolaterally convex. The shaft narrows distally. The inferior surface of the shaft has strongly developed ridges for the flexor sheaths that run the entire length of the shaft. The distal end has a well-developed but slender apical tuft. It differs from the terminal phalanges in Papio in being relatively longer, with a narrower apical tuft and less well-developed ridges for the flexor sheaths. By contrast, the terminal phalanges of Colobus are longer and more slender than in EP 902/05, with a narrower apical tuft, and more strongly developed flexor sheath ridges. An index of proximal breadth x100/maximum length of terminal phalanges (pedal and manual lateral digits combined) discriminates extant cercopithecids according to their substrate preference. Arboreal cercopithecids have relatively longer terminal phalanges (mean index: Presbytis spp. 44.9; Colobus guereza, 46.4; Cercopithecus mitis, 47.5) than terrestrial cercopithecids (Theropithecus gelada, 52.0; Papio anubis, 58.0; Erythrocebus, 63.2), with semi-terrestrial taxa broadly intermediate (Macaca nemestrina, 48.7; Mandrillus sphinx, 50.1; Chlorocebus aethiops, 55.1). The value for this index in EP 902/05 (53.0) falls into the ranges of all three categories, but it is most consistent with the breadth-length proportions of semi-terrestrial cercopithecids.
Temporal and Spatial Distribution of Fossil Cercopithecids at Laetoli As noted in the “Materials and Methods” section, few cercopithecids have been recovered from the Lower Laetolil Beds (n = 1) and Upper Ndolanya beds (n = 4). The apparent rarity of cercopithecids from the Lower Laetolil Beds (>3.8 Ma) can probably be explained as a consequence of the relative paucity of vertebrate fossils from this stratigraphic unit. The same cannot be said for the Upper Ndolanya Beds (~2.6–2.7 Ma), which have yielded abundant fossils and a diverse fauna. Cercopithecids are clearly a rare component of the Upper Ndolanya assemblage, and this may be a true
T. Harrison
reflection of their relative abundance in the fauna during the late Pliocene. Different taphonomic variables operating in the Upper Ndolanya Beds compared with the Upper Laetolil Beds (i.e., the Upper Ndolanya fauna is much more taxonomically biased against small and medium-sized mammals) may have been a contributing factor, but it is unlikely to provide a full explanation for the difference (Su and Harrison 2008). Contrasting ecological settings, with the Upper Ndolanya Beds inferred to be more arid and less tree covered than the Upper Laetolil Beds (Kingston and Harrison 2007; Kovarovic and Andrews 2007; Su and Harrison 2008), may have been a more important factor in determining the relative abundance of cercopithecids. It may also be significant that all of the Upper Ndolanya cercopithecids are known from Loc. 7E, despite the fact that large samples of fossil vertebrates from these beds have been recovered from other localities, such as Locs. 14 and 18. The cercopithecids from the Upper Ndolanya Beds can be referred to Pp. ado and cf. Rhinocolobus sp., and based on the very limited samples available, they appear to be metrically and morphologically indistinguishable from those from the Upper Laetolil Beds. Cercopithecids have been recovered from all of the main collecting localities at Laetoli, with the exception of Locs. 4, 15, 19 and 22E. This is presumably an issue of sampling, since cercopithecids only comprise 0.75% of the total mammalian assemblage (based on the Harrison collections), and these localities have yielded relatively small collections of fossil mammals (fewer than 250 specimens). In fact, most localities have cercopithecid counts that are close to the average representation relative to the total number of fossil mammals collected. A few localities (i.e., Locs. 9, 10E, 16, 21 and 22) have a significant over-representation of cercopithecids, and this may imply some ecological or taphonomic heterogeneity in the Upper Laetolil Beds. More importantly, perhaps, is the observation that cercopithecids are significantly under-represented in the lowermost horizons of the Upper Laetolil Beds below Tuff 5. Only two cercopithecid teeth were recovered by the author from horizons below Tuff 5 (i.e., at Locs. 5, 9S, 10, and 10W), whereas the expected count at these sites, based on the average representation of cercopithecids in the total mammalian samples from the Upper Laetolil Beds, should be 19. The rarity of cercopithecids in the Upper Laetolil Beds below Tuff 5 is confirmed when the samples from the Leakey and Harrison collections are combined, since only 4.3% (n = 10) of the total cercopithecid sample comes from below Tuff 5, compared with over 20% of all mammals. Clearly, cercopithecids are relatively and absolutely rarer in the earlier part of the sequence of the Upper Laetolil Beds, and this presumably relates to an important ecological difference. Paradoxically, however, the evidence from the fauna and sedimentology indicates that the paleoecology was more mesic and probably more densely wooded below Tuff 5 compared to the upper part of the
6 Laetoli Cercopithecids
sequence, and this should presumably have provided ideal habitats for cercopithecids. As noted above there is apparently a significant metrical trend in the cercopithecid samples from the Upper Laetolil Beds through time. Although the number of specimens from the early part of the stratigraphic sequence is relatively small, there does appear to be a significant increase in the size of the teeth in Pp. ado above Tuff 5 (see above for details). This may imply an evolutionary change in this lineage over the ~200 kyr duration of the deposition of the Upper Laetolil Beds. The samples are too small to examine temporal trends in cf. Rhinocolobus sp., but the single lower molar from below Tuff 5 is larger than all of the corresponding teeth from above Tuff 5. Parapapio ado is the most common taxon at Laetoli, comprising almost two thirds (65.6 %) of the cercopithecid specimens. Rhinocolobus sp. is the next most common (28.8%), followed by Cercopithecoides (4.2%) and Papionini gen. et sp. indet. (1.4%) (Table 6.1). The small sample sizes do not allow an assessment of the representation of the different species at the various localities, but the two most common species appear to have co-occurred in approximately the same proportions (Parapapio: cf. Rhinocolobus = 69:31) at different localities throughout the Upper Laetolil Beds. When compared with cercopithecid faunas from other Plio-Pleistocene localities in East Africa, there are two important observations that can be made about the composition of the Laetoli community. First is the absence of Theropithecus, which occurs as a common and ubiquitous species at Pliocene localities in East Africa.1 Second is the relatively large proportion of colobines in the fauna. As has been noted previously, these two observations are linked, since it is the success of Theropithecus in the mid-Pliocene, from ~3.4 Ma onwards, that leads to the much greater dominance of papionins and the relative decline in the proportion of colobines (Frost 2007). Although the earliest recorded occurrence of Theropithecus is ~3.9 Ma, it does not become widely distributed in East Africa until ~3.4 Ma (Frost 2007; Leakey et al. 2008). Data on the relative abundance of colobines (in terms of specimen counts) in the late Miocene and Pliocene cercopithecid faunas from East Africa, show that colobines were an important component of the communities until around ~3.5 Ma (Table 6.12). From 3.5 Ma to 1.5 Ma, during the so-called “Theropithecus zone”, 1 Ndessokia (1990: 175, 177) added Theropithecus darti to the faunal list of the Upper Laetolil Beds, based on material from his 1987–1988 collections. The location of these collections is unknown, so the identification cannot be verified. Given the absence of Theropithecus from the extensive Leakey and Harrison collections, it seems unlikely that this taxon was present in the Upper Laetolil Beds. It is possible that the material on which the identification is based is intrusive from the Ngaloba Beds, in which Theropithecus is known to occur (Harrison, personal observation).
135 Table 6.12 Relative proportion of colobines in the cercopithecid faunas from East African localities Age Locality % Colobines ~6.5–7.4 Ma ~6 Ma ~5.7 Ma ~5.2 Ma ~5.0–6.5 Ma ~4.4 Ma ~4.3–4.5 Ma ~4.1–4.4 Ma ~3.8–4.4 Ma ~4.1–4.2 Ma ~3.5–4.0 Ma
Lower Nawata Mb., Lothagam, Kenya 21 Lemudong’o, Kenya 100 Adu-Asa Fm., Ethiopia 87 Kuseralee Mb., Sagantole Fm., Ethiopia 31 Upper Nawata Mb., Lothagam, Kenya 36 Aramis, Ethiopia 56 As Duma, Gona, Ethiopia 35 Lonyumun Mb., Nachukui Fm., Kenya 31 Asa Issie, Ethiopia 86 Kanapoi, Kenya 23 28 Upper Lonyumun, Moiti and Lower Lokochot Mb., Koobi Fora Fm., Kenya ~3.5–3.8 Ma Upper Laetolil Beds, Laetoli, 33 Tanzania
~3.5–3.8 Ma Am-Ado, Ethiopia 9 ~3.5 Ma Kaiyumung Mb., Nachukui Fm., Kenya 7 ~3.3–3.4 Ma Sidi Hakoma Mb., Hadar Fm., Ethiopia 3 ~3.2–3.3 Ma Denan Dora Mb., Hadar Fm., Ethiopia 0 ~2.5–3.5 Ma Upper Lokochot and lower Burgi Mb., 16 Koobi Fora Fm., Kenya ~2.5–3.4 Ma Omo Shungura Mb. B-C, Ethiopia 5 ~1.9–2.4 Ma Omo Shungura Mb. E-G, Ethiopia 7 ~1.6–2.0 Ma Upper Burgi to KBS Mb., Koobi Fora 18 Fm., Kenya The line in the middle of the Table separates those localities (bottom of Table) with Theropithecus Data sources: Eck 1976; Harris et al. 1988; Frost 2001, 2007; Feibel 2003; Leakey et al. 2003; Leakey et al. 2008; McDougall and Feibel 2003; Semaw et al. 2005; White et al. 2006; Hlusko 2007; Reed 2008; McDougall and Brown 2008; Frost et al. 2009; Haile-Selassie et al. 2010; Harrison, unpublished
Theropithecus becomes the dominant cercopithecid at all localities, and the proportion of colobines declines accordingly (Frost 2007; Leakey et al. 2008). The Upper Laetolil Beds sample the time period (~3.5–3.8 Ma), just preceding the appearance of Theropithecus at localities throughout East Africa, when colobines and Parapapio were still the dominant cercopithecids. However, Haile-Selassie et al. (2010) have recently reported that the cercopithecid fauna from Woranso-Mille, dating to 3.6–3.8 Ma, and contemporary with the Upper Laetolil fauna, was dominated by Theropithecus oswaldi aff. darti (see Am-Ado in Table 6.12). This might imply that there were important regional differences in the turnover of the cercopithecid fauna during the mid-Pliocene. Although cercopithecids are rare in the Upper Ndolanya Beds (~2.6–2.7 Ma), it is interesting that Theropithecus has not yet been recorded from this unit, even though it is very common at other late Pliocene localities in East Africa. Therefore, the absence of Theropithecus at Laetoli might not be simply a matter of time, but may reflect the lack of suitable habitats, including a requirement for closed woodland and/or permanent bodies of water.
136
Conclusions New fossil cercopithecids recovered from Laetoli since 1998 have increased the available sample to 237 specimens. Most of the material consists of isolated teeth, jaw fragments and postcranial bones from the Upper Laetolil Beds (~3.5–3.8 Ma), but four specimens are known from the Upper Ndolanya Beds (~2.6–2.7 Ma) and a proximal humerus has been recovered from the Lower Laetolil Beds (~3.8–4.3 Ma). The enlarged sample now available for study has helped to clarify the taxonomic affinities and paleobiology of the Laetoli cercopithecids. However, it has not been possible to assign most of the taxa to known species because of the lack of relatively complete cranial material and because of the fact that the Laetoli taxa appear to be unrecorded or poorly represented at other African Pliocene localities. Four species are represented at Laetoli: Parapapio ado, Papionini gen.et sp. indet., cf. Rhinocolobus sp., and Cercopithecoides sp. Parapapio ado is the most common species at Laetoli, representing about two-thirds of the specimens. Comparisons of the dentition, mandible and face confirm that Pp. ado shares a suite of primitive features with other species of Parapapio, which distinguish it from all other genera of extant and extinct papionins. In terms of dental size and proportions, and features of the face, Pp. ado can be distinguished from all other species of Parapapio from the Plio-Pleistocene of East and South Africa. Parapapio lothagamensis and material referred to Pp. ado from Kanapoi and Allia Bay in Kenya appear to be morphologically quite distinct from other species of Parapapio, and probably represent a distinct genus. In this case, Pp. ado from Laetoli would represent the earliest record of the genus. Other than Laetoli, the only confirmed record of Pp. ado is from the contemporary Kaiyumung Mb. at Lothagam in northern Kenya (Leakey et al. 2003; Harrison, unpublished observation). The postcranial specimens attributed to Pp. ado indicate that it was a relatively slender and agile semi-terrestrial monkey (generally similar, in terms of positional behavior, to Cercocebus and some species of Macaca), adept in the trees, but frequently traveled on the ground (as confirmed by the fossil footprint evidence). A second species of papionin, larger in dental size than Pp. ado, is represented at Laetoli by several isolated teeth. The teeth are similar in morphology to extant Papio sp., but are larger or fall in the upper end of the range of variation. They are most comparable in size to Dinopithecus. Unfortunately, the material is too fragmentary and too poorly represented to establish its precise taxonomic identity. It could belong to Dinopithecus, to a large species of Papio, or to a previously unrecorded species of large papionin. The material is left unassigned as Papionini gen. et sp. indet. A distal humerus attributed to this species indicates that it was large terrestrial
T. Harrison
cercopithecid, probably less specialized for terrestriality than extant Papio. The colobines attributed to cf. Rhinocolobus sp. comprise 28.8% of the cranio-dental specimens from Laetoli. The lack of relatively complete skulls or crania hampers comparisons with other fossil colobines, and prevents a more precise taxonomic designation. Nevertheless, the material does provide adequate evidence to determine that the material does not belong to any of the previously recognized fossil colobine species from Africa, but without more complete cranial specimens it is not possible to diagnose a new species. In many aspects of the dentition, lower face and mandible it matches well with Rhinocolobus turkanaensis, but differs in having a shallower subnasal clivus, a less robust frontal process of the zygoma, a shorter rostrum, a maxillary sinus, a slightly shallower mandibular corpus, and relatively shorter upper and lower molars. With the recovery of more complete material, it is possible that this taxon could be assigned to a new genus. Although the postcranial material is insufficient to reconstruct its locomotor behavior, the large colobine from Laetoli was specialized for arboreal quadrupedalism, similar to Rhinocolobus turkanaensis, Paracolobus chemeroni and Paracolobus enkorikae. The somewhat smaller species of colobine from Laetoli was initially attributed to Colobinae gen. et sp. indet. (Leakey and Delson 1987). With the recovery of additional specimens since 1998, including a partial mandible, the taxonomic affinities of this species have been clarified. Specialized features of the dentition and mandible serve to link it to Cercopithecoides and to distinguish it from all other PlioPleistocene colobines. However, the Laetoli material has a unique set of characters that distinguish it from all of the currently recognized species of Cercopithecoides, and it certainly represents a new species. Unfortunately, the current material is not adequate to diagnose a new species, so it is recognized here as Cercopithecoides sp. The postcranial specimens attributed to this species are very similar to those of extant Colobus, and clearly indicate that it was fully arboreal in its locomotor behavior. If this is the case, Cercopithecoides sp. from Laetoli contrasts with later species of this genus, which all appear to have had postcrania that indicate a relatively high degree of terrestriality. Analysis of the distribution of fossil cercopithecids at Laetoli provides some provisional evidence of spatial patterning and temporal trends. Taking into account sampling differences, it appears that the two most common species of cercopithecids at Laetoli, Pp. ado and cf. Rhinocolobus sp., are represented in approximately the same relative proportion (ratio of 69:31) across the individual collecting localities. There is also evidence that cercopithecids are relatively underrepresented components of the mammalian fauna below Tuff 5 compared with later in the Upper Laetolil sequence. This presumably relates to key ecological differences through
6 Laetoli Cercopithecids
time during the deposition of the Upper Laetolil Beds. However, the evidence available indicates that the paleoecology below Tuff 5 was possibly slightly more mesic and more densely wooded than above Tuff 5, and this would presumably have provided equally suitable habitats for cercopithecids (Tattersfield 2011). Although the samples are relatively small, there is apparently a significant increase in size of the dentition of Pp. ado above Tuff 5, implying an evolutionary trend in this lineage over the course of the ~200 kyrs represented by the Upper Laetolil Beds. Compared with the faunas from other Plio-Pleistocene localities in East Africa, two important and related observations can be made about the composition of the Laetoli cercopithecid community: the absence of Theropithecus and the relatively large proportion of colobines in the fauna. Colobines represent an important component of cercopithecid communities from East Africa during the late Miocene and early Pliocene. From 3.5 to 1.5 Ma, Theropithecus becomes the dominant cercopithecid at all localities, and the proportion of colobines declines accordingly. The Upper Laetolil Beds sample the time period, just preceding the appearance of Theropithecus at localities throughout East Africa, when colobines and Parapapio were still the dominant cercopithecids. Acknowledgements The author is grateful to the Tanzania Commission for Science and Technology and the Unit of Antiquities in Dar es Salaam for permission to conduct research in Tanzania. Special thanks go to Paul Msemwa (Director) and Amandus Kweka, as well as to all of the staff at the National Museum of Tanzania in Dar es Salaam, for their support and assistance. The Government of Kenya and the National Museums of Kenya are thanked for permission to study the collections in Nairobi. Thanks go to Emma Mbua, Mary Muungu, Meave Leakey (Kenya National Museum), Jerry Hooker, Peter Andrews, Paula Jenkins, Daphne Hills (Natural History Museum, London), Oliver Hampe, WolfDieter Heinrich (Humboldt-Universität Museum für Naturkunde, Berlin), Nancy Simmons, Ross MacPhee, Eric Delson, and Eileen Westwig (American Museum of Natural History, New York) for access to specimens in their care. For their advice, discussion, help and support I gratefully acknowledge the following individuals: P. Andrews, H. Carter-Menn, E. Delson, S. Elton, S. Frost, C. Gilbert, L. Hlusko, N. Jablonski, C. Jolly, D.M.K. Kamamba, M.G. Leakey, C.S. Msuya, S. Odunga, M. Pickford, K. Reed, M. Rose and D. Su. I am especially grateful to S. Frost and N. Jablonski for their excellent feedback on the manuscript. Research on the Laetoli cercopithecids was supported by grants from the National Geographic Society, the Leakey Foundation, and NSF (Grants BCS-9903434 and BCS-0309513).
References Aimi, M., & Nogami, Y. (1989). Macaca fuscata develops thin enamel on lingual sides of lower incisors. Primates, 30, 261–264. Benefit, B. R., McCrossin, M., Boaz, N. T., & Pavlakis, P. (2008). New fossil cercopithecoids from the late Miocene of as Sahabi, Libya. Garyounis Scientific Bulletin, Special Issue 5, 265–282. Berger, M. E. (1972). Live-weights and body measurements of olive baboons (Papio anubis) in the Laikipia District of Kenya. Journal of Mammalogy, 53, 404–407.
137 Birchette, M. G. (1982). The postcranial skeleton of Paracolobus chemeroni. Ph.D. dissertation, Harvard University, Cambridge. Bird, M. E. (1987). A preliminary comparative study of fossil and extant mammalian tracks and trails. In M. D. Leakey & J. M. Harris (Eds.), Laetoli: A Pliocene site in northern Tanzania (pp. 546–550). Oxford: Clarendon. Bromage, T. G., & Schrenk, F. (1986). A cercopithecoid tooth from the Pliocene of Malawi. Journal of Human Evolution, 15, 497–500. Bromage, T. G., Schrenk, F., & Juwayeyi, Y. M. (1995). Paleobiogeography of the Malawi rift: Age and vertebrate paleontology of the Chiwondo beds, northern Malawi. Journal of Human Evolution, 28, 37–57. Broom, R. (1940). The South African Pleistocene cercopithecid apes. Annals of the Transvaal Museum, 20, 89–100. Ciochon, R. L. (1993). Evolution of the cercopithecoid forelimb: Phylogenetic and functional implications from morphometric analyses. University of California Publications in Geological Sciences (Vol. 138). Berkeley: University of California Press. Conroy, G. C., & Rose, M. D. (1983). The evolution of the primate foot from the earliest primates to the Miocene hominoids. Foot & Ankle, 3, 342–364. Delson, E. (1975). Evolutionary history of the Cercopithecidae. In F. S. Szalay (Ed.), Approaches to primate paleobiology. Contributions to primatology (Vol. 5, pp. 167–217). Basel: Karger. Delson, E. (1978). New fossil cercopithecids from East Africa. American Journal of Physical Anthropology, 49, 389. Delson, E. (1984). Cercopithecid biochronology of the African PlioPleistocene: Correlation among eastern and southern hominidbearing localities. Courier Forschungsinstitut Senckenberg, 69, 199–218. Delson, E., Terranova, C. J., Jungers, W. L., Sargis, E. J., Jablonski, N. G., & Dechow, P. C. (2000). Body mass in Cercopithecidae (Primates, Mammalia): Estimation and scaling in extinct and extant taxa. Anthropological Papers of the American Museum of Natural History, 83, 1–159. Dietrich, W. O. (1942). Altestquartäre Säugetiere aus der Südlichen Serengeti, Deutsch-Ostafrika. Palaeontographica, 94(A), 43–133. Eck, G. G. (1976). Cercopithecoidea from Omo Group deposits. In Y. Coppens, F. C. Howell, G. L. Isaac, & R. E. F. Leakey (Eds.), Earliest man and environments in the Lake Rudolf Basin (pp. 332– 344). Chicago: University of Chicago Press. Eck, G. (1977). Diversity and frequency distribution of Omo Group Cercopithecoidea. Journal of Human Evolution, 6, 55–63. Eisenhart, W. L. (1974). The fossil cercopithecoids of Makapansgat and Sterkfontein. B. A. thesis, Harvard College, Cambridge. Elton, S. (2001). Locomotor and habitat classifications of cercopithecoid postcranial material from Sterkfontein Member 4, Bolt’s farm and Swartkrans Member 1 and 2, South Africa. Palaeontologia Africana, 37, 115–126. El-Zaatari, S., Frine, F. E., Teaford, M. F., & Smith, H. F. (2005). Molar microwear and dietary reconstructions of fossil Cercopithecoidea from the Plio-Pleistocene deposits of South Africa. Journal of Human Evolution, 49, 180–205. Feibel, C. S. (2003). Stratigraphy and depositional setting of the Pliocene Kanapoi Formation, Lower Kerio Valley, Kenya. In J. M. Harris & M. G. Leakey (Eds.), Geology and vertebrate paleontology of the early Pliocene site of Kanapoi, northern Kenya (pp. 9–20). Los Angeles: Contributions in Science 498, Natural History Museum of Los Angeles County. Fleagle, J. G. (1977). Locomotor behavior and skeletal anatomy of sympatric Malaysian leaf-monkeys (Presbytis obscura and Presbytis melalophos). Yearbook of Physical Anthropology, 20, 440–453. Fleagle, J. G., & Meldrum, D. J. (1988). Locomotor behavior and skeletal morphology of two sympatric pitheciiine monkeys, Pithecia pithecia and Chiropotes satanus. American Journal of Physical Anthropology, 16, 227–249.
138 Freedman, L. (1957). The fossil Cercopithecoidea of South Africa. Annals of the Transvaal Museum, 23, 121–262. Freedman, L. (1976). South African fossil Cercopithecoidea: A reassessment including a description of new material from Makapansgat, Sterkfontein and Taung. Journal of Human Evolution, 5, 297–315. Freedman, L., & Stenhouse, N. S. (1972). The Parapapio species of Sterkfontein, Transvaal, South Africa. Palaeontologia Africana, 14, 93–111. Frost, S. R. (2001). New early Pliocene Cercopithecidae (Mammalia: Primates) from Aramis, Middle Awash Valley, Ethiopia. American Museum Novitates, 3350, 1–36. Frost, S. R. (2007). African Pliocene and Pleistocene cercopithecid evolution and global climatic change. In R. Bobe, Z. Alemseged, & A. K. Behrensmeyer (Eds.), Hominin environments in the East African Pliocene: An assessment of the faunal evidence (pp. 51–76). Dordrecht: Springer. Frost, S. R., & Delson, E. (2002). Fossil Cercopithecidae from the Hadar formation and surrounding areas of the Afar Depression, Ethiopia. Journal of Human Evolution, 43, 687–748. Frost, S. R., & Kullmer, O. (2008). Cercopithecidae from the Pliocene Chiwondo Beds, Malawi-rift. Geobios, 41, 743–749. Frost, S. R., Plummer, T., Bishop, L. C., Ditchfield, P., Ferraro, J., & Hicks, J. (2003). Partial cranium of Cercopithecoides kimeui Leakey, 1982 from Rawi Gully, Southwestern Kenya. American Journal of Physical Anthropology, 122, 191–199. Frost, S. R., Haile-Selassie, Y., & Hlusko, L. (2009). Cercopithecidae. In Y. Haile-Selassie & G. WoldeGabriel (Eds.), Ardipithecus kadabba: Late Miocene evidence from the Middle Awash Ethiopia (pp. 135–158). Berkeley: University of California Press. Gilbert, C. C. (2007). Craniomandibular morphology supporting the diphyletic origin of mangabeys and a new genus of the Cercocebus/Mandrillus clade, Procercocebus. Journal of Human Evolution, 53, 69–102. Gilbert, C. C. (2008). African papioinin phylogenetic history and PlioPleistocene biogeography. Ph.D. dissertation, Stony Brook University, Stony Brook. Gilbert, W. H., & Frost, S. (2008). Cercopithecidae. In W. H. Gilbert & B. Asfaw (Eds.), Homo erectus: Pleistocene evidence from the Middle Awash, Ethiopia (pp. 115–132). Berkeley: University of California Press. Grine, F. E., & Hendey, Q. B. (1981). Earliest primate remains from South Africa. South African Journal of Science, 77, 374–376. Haile-Selassie, Y., WoldeGabriel, G., White, T. D., Bernor, R. L., Degusta, D., Renne, P. R., Hart, W. K., Vrba, E., Ambrose, S., & Howell, F. C. (2004). Mio-Pliocene mammals from the Middle Awash, Ethiopia. Geobios, 37, 536–552. Haile-Selassie, Y., Saylor, B. Z., Deino, A., Alene, M., & Latimer, B. M. (2010). New hominid fossils from Woranso-Mille (Central Afar, Ethiopia) and taxonomy of early Australopithecus. American Journal of Physical Anthropology, 141, 406–417. Harris, J. M., Brown, F. H., & Leakey, M. G. (1988). Stratigraphy and paleontology of Pliocene and Pleistocene localities west of Lake Turkana, Kenya. Contributions in Science, 399, 1–128. Harris, J. M., Leakey, M. G., & Cerling, T. E. (2003). Early Pliocene tetrapod remains from Kanapoi, Lake Turkana Basin, Kenya. In J. M. Harris & M. G. Leakey (Eds.), Geology and vertebrate paleontology of the early Pliocene site of Kanapoi, northern Kenya (pp. 39–113). Los Angeles: Contributions in Science 498, Natural History Museum of Los Angeles County. Harrison, T. (1982). Small-bodied apes from the Miocene of East Africa. Ph.D. dissertation, University of London, London. Harrison, T. (1989). New postcranial remains of Victoriapithecus from the middle Miocene of Kenya. Journal of Human Evolution, 18, 3–54. Harrison, T., & Harris, E. E. (1996). Plio-Pleistocene cercopithecids from Kanam East, Western Kenya. Journal of Human Evolution, 30, 539–561.
T. Harrison Harrison, T., & Kweka, A. (2011). Paleontological localities. In T. Harrison (Ed.), Paleontology and geology of Laetoli: Human evolution in context (Geology, geochronology, paleoecology and paleoenvironment, Vol. 1, pp. 17–45). Dordrecht: Springer. Haughton, S. H. (1925). A note on the occurrence of a species of baboon in limestone deposits near Taung. Transactions of the Royal Society of South Africa, 12, lxviii. Heaton, J. L. (2006). Taxonomy of the Sterkfontein fossil Cercopith ecinae: The Papionini of Members 2 and 4 (Gauteng, South Africa). Ph.D. dissertation, Indiana University, Bloomington. Hlusko, L. J. (2006). A new large Pliocene colobine species (Mammalia: Primates) from Asa Issie, Ethiopia. Geobios, 39, 57–69. Hlusko, L. J. (2007). A new late Miocene species of Paracolobus and other Cercopithecoidea (Mammalia: Primates) fossils from Lemudong’o, Kenya. Kirtlandia, 56, 72–85. Hopwood, A. T. (1936). New and little-known fossil mammals from the Pleistocene of Kenya colony and Tanganyika Territory. Annals and Magazine of Natural History, Ser., 10(17), 636–638. Howell, A. B., & Straus, W. L. (1933). The muscular system. In C. G. Hartman & W. L. Straus (Eds.), The anatomy of the rhesus monkey (Macaca mulatta) (pp. 89–175). New York: Hafner. Iwamoto, M. (1982). A fossil baboon skull from the Lower Omo Basin, southwestern Ethiopia. Primates, 23, 533–541. Jablonski, N. G. (1994). New fossil cercopithecid remains from the Humpata Plateau, southern Angola. American Journal of Physical Anthropology, 94, 435–464. Jablonski, N. G. (2002). Fossil Old World monkeys: The late Neogene radiation. In W. C. Hartwig (Ed.), The primate fossil record (pp. 255–299). Cambridge: Cambridge University Press. Jablonski, N. G., Leakey, M. G., & Antón, M. (2008a). Systematic paleontology of the cercopithecines. In N. G. Jablonski & M. G. Leakey (Eds.), Koobi Fora research project (The fossil monkeys, Vol. 6, pp. 103–300). San Francisco: California Academy of Science. Jablonski, N. G., Leakey, M. G., Ward, C. V., & Antón, M. (2008b). Systematic paleontology of the large colobines. In N. G. Jablonski & M. G. Leakey (Eds.), Koobi Fora research project (The fossil monkeys, Vol. 6, pp. 31–102). San Francisco: California Academy of Science. Jolly, C. J. (1967). The fossil history of the baboons. In H. Vagtborg (Ed.), The baboons in medical research (pp. 23–50). Austin: University of Texas Press. Jolly, C. J. (1972). The classification and natural history of Theropithecus (Simopithecus) (Andrews, 1916), baboons of the African PlioPleistocene. Bulletin of the British Museum (Natural History) Geology, 22(1), 1–123. Jones, T. R. (1937). A new fossil primate from Sterkfontein, Krugersdorp, Transvaal. South African Journal of Science, 33, 709–728. Kingston, J. D., & Harrison, T. (2007). Isotopic dietary reconstructions of Pliocene herbivores at Laetoli: Implications for hominin paleoecology. Palaeogeography, Palaeoclimatology, Palaeoecology, 243, 272–306. Koppe, T., & Swindler, D. R. (2004). Metrical sexual dimorphism in the deciduous teeth of Old World monkeys. Annals of Anatomy, 186, 367–374. Kovarovic, K., & Andrews, P. (2007). Bovid postcranial ecomorphological survey of the Laetoli paleoenvironment. Journal of Human Evolution, 52, 663–680. Larson, S. G. (1993). Functional morphology of the shoulder in primates. In D. L. Gebo (Ed.), Postcranial adaptation in nonhuman primates (pp. 45–69). De Kalb: Northern Illinois University Press. Leakey, M. G. (1982). Extinct large colobines from the Plio-Pleistocene of Africa. American Journal of Physical Anthropology, 58, 153–172. Leakey, M. D. (1987a). Introduction. In M. D. Leakey & J. M. Harris (Eds.), Laetoli: A Pliocene site in northern Tanzania (pp. 1–22). Oxford: Clarendon. Leakey, M. D. (1987b). Animal prints and trails. In M. D. Leakey & J. M. Harris (Eds.), Laetoli: A Pliocene site in northern Tanzania (pp. 451–489). Oxford: Clarendon.
6 Laetoli Cercopithecids Leakey, M. G. (1987c). Colobinae (Mammalia, Primates) from the Omo Valley, Ethiopia. In Y. Coppens & F. C. Howell (Eds.), Les Faunes Plio-Pléistocenès de la Basse Vallée de l’Omo (Éthiopie). Tome 3, Cercopithecidae de la Formation de Shungura. Cahiers de Paléontologie (pp. 149–169). Paris: CNRS. Leakey, M. G., & Delson, E. (1987). Fossil Cercopithecidae from the Laetolil Beds. In M. D. Leakey & J. M. Harris (Eds.), Laetoli: A Pliocene site in northern Tanzania (pp. 91–107). Oxford: Clarendon. Leakey, M. G., & Leakey, R. E. F. (1976). Further Cercopithecinae (Mammalia, Primates) from the Plio/Pleistocene of East Africa. Fossil Vertebrates of Africa, 4, 121–146. Leakey, M. G., Feibel, C. S., McDougall, I., & Walker, A. (1995). New four-million-year-old hominid species from Kanapoi and Allia Bay, Kenya. Nature, 376, 565–571. Leakey, M. G., Teaford, M. F., & Ward, C. V. (2003). Cercopithecidae from Lothagam. In M. G. Leakey & J. M. Harris (Eds.), Lothagam: The dawn of humanity in eastern Africa (pp. 204–248). New York: Columbia University Press. Leakey, M. G., Gathogo, P., & Jablonski, N. G. (2008). Geological background and cercopithecoid faunal assemblages. In N. G. Jablonski & M. G. Leakey (Eds.), Koobi Fora research project (The fossil monkeys, Vol. 6, pp. 359–396). San Francisco: California Academy of Science. Le Minor, J. M. (1987). Comparative anatomy and significance of the sesamoid bone of the peroneus longus muscle (os peroneum). Journal of Anatomy, 151, 85–99. McDougall, I., & Brown, F. H. (2008). Geochronology of the pre-KBS Tuff sequence, Omo Group, Turkana Basin. Journal of the Geological Society of London, 165, 549–562. McDougall, I., & Feibel, C. S. (2003). Numerical age control for the Miocene-Pliocene succession at Lothagam, a hominoid-bearing sequence in the northern Kenya Rift. In M. G. Leakey & J. M. Harris (Eds.), Lothagam: The dawn of humanity in eastern Africa (pp. 43–60). New York: Columbia University Press. Mollett, O. (1947). Fossil mammals from the Makapansgat Valley. Potgeitersrust. I. Primates. South African Journal of Science, 43, 95–303. Ndessokia, P. N. S. (1990). The mammalian fauna and archaeology of the Ndolanya and Olpiro Beds, Laetoli, Tanzania. Ph.D. dissertation, University of California, Berkeley. Patterson, B. (1968). The extinct baboon, Parapapio jonesi, in the early Pleistocene of northwestern Kenya. Breviora, 282, 1–4. Pickford, M., Mein, P., & Senut, B. (1992). Primate bearing PlioPleistocene cave deposits of Humpata, southern Angola. Human Evolution, 7, 17–33. Rae, T. C. (2008). Paranasal pneumatization in extant and fossil Cercopithecoidea. Journal of Human Evolution, 54, 279–286. Reck, H., & Kohl-Larsen, L. (1936). Erster Überblick über die jungdiluvialen Tier- und Menschenfunde Dr. Kohl-Larsen’s im nordöstlichen Teil des Njarasa-Grabens (Ostafrika) und die geologischen Verhältnisse des Fundgebietes. Geologische Rundschau, 27, 401–441.
139 Reed, K. E. (2008). Paleoecological patterns at the Hadar hominin site, Afar Regional State, Ethiopia. Journal of Human Evolution, 54, 743–768. Rose, M. D. (1989). New postcranial specimens of catarrhines from the Middle Miocene Chinji Formation, Pakistan: Descriptions and a discussion of proximal humerus functional morphology in anthropoids. Journal of Human Evolution, 18, 131–162. Rose, M. D. (1993). Functional anatomy of the elbow and forearm in primates. In D. L. Gebo (Ed.), Postcranial adaptation in nonhuman primates (pp. 70–95). De Kalb: Northern Illinois University Press. Semaw, S., Simpson, S. W., Quade, J., Renne, P. R., Butler, R. F., McIntosh, W. C., Levin, N., Dominguez-Rodrigo, M., & Rogers, M. J. (2005). Early Pliocene hominids from Gona, Ethiopia. Nature, 433, 301–305. Senut, B. (1994). Cercopithecoidea Néogènes et Quaternaires du Rift Occidental (Ouganda). In B. Senut & M. Pickford (Eds.), Geology and palaeobiology of the Albertine Rift Valley, Uganda-Zaire (Palaeobiology, Vol. II, pp. 195–205). Orléans: CIFEG. Senut, B. (1996). Cercopithecoidea du Plio-Pléistocène des Koanaka Hills (Ngamiland, Botswana). Comptes rendus de l’Académie des Sciences Paris, 322, Séries IIa, 423–428. Smith, J. M., & Savage, R. J. G. (1956). Some locomotory adaptations in mammals. Journal of the Linnean Society (Zoology), 42, 603–622. Strasser, E. (1988). Pedal evidence for the origin and diversification of cercopithecid clades. Journal of Human Evolution, 17, 225–245. Strasser, E., & Delson, E. (1987). Cladistic analysis of cercopithecid relationships. Journal of Human Evolution, 16, 81–99. Su, D. F., & Harrison, T. (2008). Ecological implications of the relative rarity of fossil hominins at Laetoli. Journal of Human Evolution, 55, 672–681. Swindler, D. R. (2002). Primate dentition: An introduction to the teeth of non-human primates. Cambridge: Cambridge University Press. Szalay, F. S., & Delson, E. (1979). Evolutionary history of the primates. New York: Academic. Tattersfield, P. (2011). Gastropoda. In T. Harrison (Ed.), Paleontology and geology of Laetoli: Human evolution in context (Fossil hominins and the associated fauna, Vol. 2, pp. 567–587). Dordrecht: Springer. Walker, A. (1974). Locomotor adaptations in past and present prosimian primates. In F. A. Jenkins (Ed.), Primate locomotion (pp. 349–381). New York: Academic. White, T. D., WoldeGabriel, G., Asfaw, B., Ambrose, S., Beyene, Y., Bernor, R. L., Boisserrie, J.-R., Currie, B., Gilbert, H., HaileSelassie, Y., Hart, W. K., Hlusko, L. J., Howell, F. C., Kono, R., Lehmann, T., Louchart, A., Lovejoy, C. O., Renne, P. R., Saegusa, H., Vrba, E. S., & Suwa, G. (2006). Asa Issie, Aramis and the origin of Australopithecus. Nature, 440, 883–889. Williams, F. L., Ackermann, R. R., & Leigh, S. R. (2007). Inferring Plio-Pleistocene southern African biochronology from facial affinities in Parapapio and other fossil papionins. American Journal of Physical Anthropology, 132, 163–174.
wwwwwwwwwwwwwww
Chapter 7
Hominins from the Upper Laetolil and Upper Ndolanya Beds, Laetoli Terry Harrison
Abstract Renewed investigations at Laetoli in northern Tanzania have led to the recovery of a number of new fossil hominins. A lower canine and a mandibular fragment from the Upper Laetolil Beds (3.63–3.85 Ma) are referred to Australopithecus afarensis, and an edentulous maxilla and a proximal tibia from the Upper Ndolanya Beds (2.66 Ma) are attributed to Paranthropus aethiopicus and Hominini gen. et sp. indet., respectively. Additional hominin specimens from earlier collections are described here for the first time, including three specimens of A. afarensis, probably from the Upper Laetolil Beds, and a possible cranial fragment of an infant from the Upper Ndolanya Beds. The chronology and provenance of the Laetoli hominins are reconsidered. The species afarensis is provisionally retained in Australopithecus to reflect its anatomical and paleobiological similarities to the other species of Australopithecus sensu lato, but a reasonable case could be made on phylogenetic grounds to transfer it to Praeanthropus. It has been argued that the Laetoli sample of A. afarensis is morphologically and temporally intermediate between A. anamensis and the Hadar sample of A. afarensis, and that A. anamensis and A. afarensis represent a single anagenetically evolving lineage. However, the new specimens from the Upper Laetolil Beds help to close the gap between the Laetoli and Hadar samples, and a critical assessment of the morphological variation in the two samples indicates that there are few consistent differences separating them. Rather than being intermediate in morphology, the Laetoli sample appears to represents an earlier population of A. afarensis, with almost the full complement of derived features that characterizes the Hadar sample, but still retaining a few primitive traits. The morphological features that distinguish A. anamensis from A. afarensis are much more extensive, and these provide adequate justification for the recognition of a species distinction. The evidence best fits an evolutionary model involving a cladogenetic event rather than a simple anagenetic transformation of a single unbranched anamensis-afarensis lineage through time. The Paranthopus aethiopicus T. Harrison (*) Center for the Study of Human Origins, Department of Anthropology New York University, 25 Waverly Place, New York, NY 10003, USA e-mail:
[email protected] specimen from the Upper Ndolanya Beds represents the oldest securely dated specimen definitively attributable to this taxon and the first definitive record outside of the Turkana Basin. The Paranthropus clade probably immigrated into eastern Africa before 2.7 Ma, and became widely distributed throughout the region soon thereafter. The timing and biogeographic patterning of the occurrence of Paranthropus and Homo suggest that their respective dispersals into eastern Africa were not coincident or synchronous. Homo appeared somewhat later than Paranthropus across most of eastern Africa, except in the Awash region of Ethiopia where Homo makes its first appearance in the absence of Paranthropus. These differences in the timing and distribution suggest that Paranthropus and Homo may have had different biogeographic histories, and that their ancestral species may have had different ecological requirements at the time of their initial influx into eastern Africa. Keywords Pliocene • Tanzania • Australopithecus afarensis • Paranthropus aethiopicus • Laetoli • Hadar • Taxonomy
Introduction Laetoli in northern Tanzania has yielded a relatively small but important collection of early hominins from the midPliocene Upper Laetolil Beds, dated from ~3.63–3.83 Ma (Weinert 1950; White 1977, 1980a, 1981; Leakey 1987a, b; Kyauka and Ndessokia 1990). Although there has been debate about the number of taxa represented at Laetoli, and what these taxa should be named (Tobias 1980a, b; Johanson 1980a; Day et al. 1980a; White et al. 1981; Olson 1981, 1985; Logan et al. 1983; Ferguson 1986, 1987, 1988; Falk et al. 1995; Groves 1996; Senut 1996; Kimbel et al. 2004; Grine et al. 2006), there is current consensus that the remains can all be attributed to a single species, Australopithecus afarensis (or Praeanthropus afarensis) (see Table 7.1). The Laetoli sample of A. afarensis (n = 33) is not as large as the collections from Hadar (330 specimens; Kimbel and Delezene 2009), and the specimens tend to be isolated elements and
T. Harrison (ed.), Paleontology and Geology of Laetoli: Human Evolution in Context. Volume 2: Fossil Hominins and the Associated Fauna, Vertebrate Paleobiology and Paleoanthropology, DOI 10.1007/978-90-481-9962-4_7, © Springer Science+Business Media B.V. 2011
141
142 Table 7.1 Taxonomy and synonymy list of Australopithecus afarensis Superfamily: Hominoidea Gray, 1825 Family: Hominidae Gray, 1825 Subfamily: Homininae Gray, 1825 Tribe: Hominini Gray, 1825 Genus: Australopithecus Dart, 1925 Species: A. afarensis Johanson, 1978 Synonymy 1948 – Praeanthropus Hennig, 1948 – Hennig (1948) [nomen nudum, no fixation of type species] 1950 – Meganthropus africanus Weinert, 1950 – Weinert (1950) 1954 – Australopithecus africanus transvaalensis Broom, 1936 – Robinson (1954) [partim] 1955 – Praeanthropus africanus (Weinert 1950) – Senyürek (1955) 1978 – Australopithecus afarensis Johanson, 1978 – Hinrichsen (1978) 1978 – Australopithecus afarensis Johanson, White and Coppens, 1978 – Johanson et al. (1978) 1980 – Australopithecus africanus afarensis Johanson, 1978 – Tobias (1980b) 1980 – Australopithecus africanus aethiopicus Tobias, 1980 – Tobias (1980b) [nomen nudum, conditionally proposed] 1980 – Australopithecus africanus tanzaniensis Tobias, 1980 – Tobias (1980b) [nomen nudum, conditionally proposed] 1981 – Paranthropus africanus (Weinert 1950) – Olson (1981) 1981 – Homo (Australopithecus) sp. indet. – Olson (1981) 1983 – Dryopithecus (Sivapithecus) sivalensis (Lydekker, 1879) – Ferguson (1983) 1984 – Homo antiquus Ferguson, 1984 – Ferguson (1984) [junior homonym, name preoccupied by Homo antiquus Adloff, 1908] 1985 – Homo (Australopithecus) aethiopicus (Tobias 1980b) – Olson (1985) [junior homonym, name preoccupied by Homo aethiopicus Bory de Saint-Vincent, 1825] 1987 – Australopithecus africanus miodentatus Ferguson, 1987 – Ferguson (1987) 1996 – Australopithecus antiquus (Ferguson 1984) – Senut (1996) 1996 – Australopithecus bahrelghazali Brunet et al., 1996 – Brunet et al. (1996) 1999 – africanus Weinert, 1950 – name suppressed by the International Commission on Zoological Nomenclature for the purposes of the Principle of Priority but not for those of the Principle of Homonymy, Opinion 1941 2000 – Praeanthropus afarensis Johanson, 1978 – Wood and Richmond (2000)
more fragmentary. However, the sample does represent the second largest sample of A. afarensis, and perhaps more significantly derives from an earlier time period (the hominins from Hadar date from ~3.0–3.4 Ma; Kimbel et al. 2004; Campisano and Feibel 2008). The fossil hominins from Woranso-Mille (~3.57–3.8 Ma) in Ethiopia, with 30 specimens recovered to date, including a partial skeleton (HaileSelassie et al. 2007, 2010), have not yet been formally taxonomically assigned, but if they later prove to belong to A. afarensis they would provide another important sample of this taxon contemporary with Laetoli. The site of Laetoli is unique for the remarkable preservation of trails of footprints, presumably of A. afarensis (see Leakey
T. Harrison
and Hay 1979; Clarke 1979; Day and Wickens 1980; White 1980b; Charteris et al. 1981, 1982; Hay and Leakey 1982; Leakey 1978, 1979, 1981, 1987c; White and Suwa 1987; Tuttle 1985, 1987, 1990, 1994, 2008; Tuttle et al. 1990, 1991a, b, 1992; Feibel et al. 1996; Agnew and Demas 1998; Meldrum 2004; Sellers et al. 2005; Berge et al. 2006; Raichlen et al. 2008). This evidence has been used to confirm earlier inferences based on functional morphology of the skeletal remains that bipedalism was an important component of the terrestrial locomotor behavior of mid-Pliocene hominins (e.g., Johanson et al. 1982; Stern and Susman 1983; Susman et al. 1984, 1985; Latimer et al. 1987; Latimer 1991; Susman and Stern 1991; McHenry 1986, 1991, 1994; Stern 2000; Ward 2002). Fossil hominins have not yet been recovered from the Lower Laetolil Beds (~4.4–3.85 Ma; Deino 2011), and are rare in the younger stratigraphic units that overlie the Upper Laetolil Beds, although indirect evidence of their presence is provided by the occurrence of stone tools in the Olpiro Beds (~2.0 Ma; Deino 2011) and Ngaloba Beds (Late Pleistocene) (Harris and Harris 1981; Leakey 1987a; Hay 1987; Ndessokia 1990). Mary Leakey’s expeditions did recover a relatively complete cranium of Homo sapiens from the Late Pleistocene Upper Ngaloba Beds. Most recently, Harrison (2002) reported specimens attributable to Paranthropus aethiopicus from the Upper Ndolanya Beds, dated to 2.66 Ma (Deino 2011) (see Fig. 7.1). The history of discovery of A. afarensis and the other hominin finds from Laetoli is briefly reviewed below.
Fig. 7.1 Stratigraphic column and radiometric dating of the lower part of the sequence at Laetoli (After Hay 1987; Drake and Curtis 1987; Ndessokia 1990; Manega 1993; Mollel et al. 2011; Deino 2011)
7 Hominins from Laetoli
The fossil collections made by Louis and Mary Leakey at Laetoli in 1935 included a hominin lower canine (M.42323, formerly M.18773), which is housed in the Natural History Museum in London. This was the first Pliocene hominin to be recovered from East Africa, although the specimen was not identified as such until some decades later (White 1981). In 1939 Kohl-Larsen’s expedition to Garusi (= Laetoli) included a hominin maxilla with P3 and P4 (Garusi I) and an isolated M3 (Garusi II) (Weinert 1950; Remane 1950, 1954; Robinson 1953, 1955; Protsch 1976, 1981). Both of these specimens are housed in the Institut für Ur- und Frühgeschichte und Archäologie des Mittelalters, Tübingen. An undescribed occipital fragment (Garusi III), apparently of a fossil hominin from Pleistocene sediments, has been lost (Protsch 1976, 1981; Ullrich 2001). In the late 1970s, Eric Delson identified a previously unrecognized hominin lower incisor among the fossil cercopithecids in Berlin that had been collected by Kohl-Larsen (White 1981; Delson, personal communication), and this specimen has been briefly described (Ullrich 2001). The most extensive collection of hominins from the Upper Laetoli Beds (n = 25) was recovered by expeditions led by Mary Leakey from 1974 to 1979 (Leakey 1987b). These comprise 14 isolated teeth, 10 cranial/jaw fragments or associated dentitions, and a partial skeleton of an immature individual. Of these, 23 have been described previously (White 1977, 1980a; Leakey 1987b), and two are identified here as belonging to A. afarensis for the first time. These include a weathered and heavily rolled mandibular fragment (L.H. 29), initially referred to Homo cf. H. erectus, and a weathered isolated upper canine (LAET 79-5447), which were found the same field season at Loc. 8. At the time of their discovery these surface finds were presumed to be derived from the deflated Pleistocene sediments, because they have the black and orange staining typical of the fossils from these beds. However, specimens that erode out of the Upper Laetolil Beds and are reworked into the superficial lag deposits can often develop similar preservational characteristics. Since their morphology is entirely consistent with material from the Upper Laetolil Beds, these two specimens are reassigned here to A. afarensis. A further undescribed specimen (LAET 75-3817) of a possible hominin was excavated by Mary Leakey’s team at Loc. 7E from the Upper Ndolanya Beds. This is a zygomatic process of a right frontal of an infant, recorded in the catalogue as a cercopithecid. The only other hominin recovered by Mary Leakey’s expeditions is a hominin cranium (L.H. 18) from the Late Pleistocene Upper Ngaloba Beds, referrable to Homo sapiens (Day et al. 1980b; Magori and Day 1983). Renewed investigations at Laetoli by the Institute of Human Origins, directed by D.C. Johanson from 1985 to 1988, succeeded in recovering a single hominin specimen,
143
an isolated right M3 (L.H. 31) (Ndessokia 1990; Kyauka and Ndessokia 1990). The specimen was recovered from the Upper Laetolil Beds at Loc. 10 in 1987, but the precise stratigraphic provenance is unknown. Unfortunately, the author has not been an able to relocate the specimen. The resumption of full-scale paleontological and geological research at Laetoli and at other sites on the Eyasi Plateau in 1998, under the direction of the author, led to the recovery of four additional hominins. Two specimens, an isolated lower canine (EP 162/00) and a mandibular fragment with P3-M1 (EP 2400/00), were recovered from the Upper Laetolil Beds at Loc. 16 (see Harrison and Kweka 2011; Fig. 7.2). Both specimens are referable to A. afarensis. In addition, two hominins were recovered from the Upper Ndolanya Beds – a proximal tibia (EP 1000/98) and an edentulous maxilla (EP 1500/01). These are the first hominins to be recovered from this stratigraphic unit, and have proven to be of exceptional interest. EP 1500/01 was found in 2001 at the newly recorded locality of Silal Artum, just to the north of the main fossiliferous outcrops at Laetoli (see Harrison and Kweka 2011; Fig. 7.2). As is discussed below, the maxilla can be confidently attributed to Paranthropus aethiopicus. EP 1000/98 was found during the first season of renewed fieldwork at Laetoli in 1998, at yet another new locality, Loc. 22S, this time at the southern edge of the main Laetoli outcrops (see Harrison and Kweka 2011; Fig. 7.2). Attribution of isolated postcranial specimens to early hominin taxa is obviously problematic, but given the occurrence of P. aethiopicus as the only hominin known from the Upper Ndolanya Beds it is likely that the proximal tibia belongs to this species. However, since there is no direct association and the hominins from the Upper Ndolanya Beds are few, the proximal tibia is conservatively identified here as Hominini gen. et sp. indet. Further discussion concerning the affinities of EP 1000/98 is presented below. The aim of this contribution is to present a descriptive account of the morphology of the newly collected hominin specimens from the Upper Laetolil and Upper Ndolanya Beds, as well as of the previously undescribed specimens from the Kohl-Larsen and Mary Leakey collections. The study also provides an opportunity to clarify aspects of the chronology and provenance of the Laetoli hominins, and to discuss their implications for understanding the evolutionary history of Australopithecus afarensis and Paranthropus aethiopicus.
Material A list of hominins recovered from Laetoli between 1935 and 1979 was presented in Leakey (1987b: 116–117), and the individual specimens have been figured and described in
144
T. Harrison
Fig. 7.2 Map of the Laetoli area showing the main outcrops of the Upper Laetolil and Upper Ndolanya Beds and the paleontological collecting localities
some detail (Leakey et al. 1976; White 1977, 1980a, 1981; Day et al. 1980b; Magori and Day 1983; Leakey 1987b). An updated list is presented in Table 7.2, which includes a number of emendations and corrections to the original list, as well as the addition of new specimens recovered or identified since 1987. All of the hominins recovered from the Upper Laetolil Beds can be referred to Australopithecus afarensis. The only taxon so far identified in the Upper Ndolanya Beds is Paranthropus aethiopicus. The Laetoli hominins are housed in the Natural History Museum in London (NHM.M; 1935 Leakey collection), Humboldt-Universität Museum für Naturkunde in Berlin (MB Ma.; 1938–1939 Kohl-Larsen collection), Institut für Ur- und Frühgeschichte und Archäologie des Mittelalters, Tübingen (Garusi hominins; 1938–1939 Kohl-Larsen collection), Kenya National Museum in Nairobi (LAET; 1974– 1979 Leakey collections on loan from Tanzania), and National Museum of Tanzania (EP, Eyasi Plateau expedition;
1998–2005 Harrison collection). Comparisons with original specimens and casts of Australopithecus afarensis, A. anamensis, Paranthropus boisei, P. robustus, and P. aethiopicus were carried out at the Kenya National Museum (KNM), National Museum of Tanzania (TNM), American Museum of Natural History (AMNH), and Natural History Museum in London (NHM).
Provenance Most of the hominins from Laetoli, including the new specimens described here, are surface finds that had already eroded out of their original stratigraphic context at the time of their discovery. In most cases, based on the position of the find spot, the local topographic and stratigraphic context, and the occurrence of associated fossils, it is possible to determine the
1939
1939
1939
1974
1974
1974–75
1974
1974 1974–75
1975 1975 1975
1975 1975 1975
1975–76
1976
1976 1976 1976
Garusi I
Garusi II
MB Ma. 8294 (Garusi 4) L.H. 1
L.H. 2
L.H. 3
L.H. 4
L.H. 5 L.H. 6
L.H. 7 L.H. 8 L.H. 10
L.H. 11 L.H. 12 L.H. 13
L.H. 14
L.H. 15
L.H. 16 L.H. 17 L.H. 18
L. Kangiran A. Mwongela E. Kandindi
Mrs. Luce
E. Kandindi
E. Kandindi E. Kandindi M. Jackes
M. Mwoka E. Kandindi E. Kandindi
M. Muluila M. Mwoka
M. Muluila
M. Mwoka
M. Muluila
Loc. 6 Loc. 9 Loc. 2
Loc. 1
Loc. 19
Loc. 10W Loc. 5 Loc. 8
Loc. 5 Loc. 11 Loc. 10W
Loc. 8 Loc. 7
Loc. 7
Loc. 7
Loc. 3
Loc. 1
Unknown
Unknown
Unknown
Just below Tuff 6 Between Tuffs 5 and 8 Upper Ngaloba Beds
0.9 m above Tuff 8
~0.3 m above Tuff 5
7.3 m below Tuff 3 1.8 m below Tuff 4 ~3 m below Tuff 7
~0.6 m above Tuff 5 ~0.9 m above Tuff 7 5.5 m below Tuff 3
~1.8 m below Tuff 7 ~0.5 m above Tuff 7
1.2 m below Tuff 7
Between Tuffs 7 and 8, just above xenolith horizon
~0.5 m above Tuff 7
~1.2 m above Tuff 7
Upper Laetolil Beds
Upper Laetolil Beds
Upper Laetolil Beds
Rt M1 Lt M1 Cranium
Mandible (infant) with rt and lt dP4, unerupted crowns of rt and lt I1, C, P3; damaged crowns of rt and lt dC, dP3 and M1 Partial upper and lower dentition: (a) rt dP4, (b) lt I1, (c) rt I2, (d) lt I2, (e) lt C1, (f) lt P3, (g) lt P4, (h) rt M1, (i) lt M1 frag., (j) lt dI2, (k) rt dC1, (l) rt dP3, (m) rt I1, (n) rt C1, (o) lt C1, (p) rt P3, (q) rt P4 (r) lt P4, (s) rt M1 frag., (t) lt M1 Mandible with rt C-M3, lt P4-M2, roots lt C-P3, alveoli rt and lt incisors Rt maxilla frag. with I2-M1 Associated upper teeth: (a) rt I2, (b) rt C1, (c) rt P3, (d) rt dP4 and unerupted P4, (e) rt M1 Rt M3 Rt M2 and M3 Lt mandibular frag., edentulous with broken roots C-M1 Lt M2 Lt M3 Rt mandibular frag., edentulous with roots of M1-M3 Associated lower dentition: (a) rt I1, (b) lt I1, (c) rt I2, (d) lt I2, (e) rt C1, (f) lt C1, (g) lt P4, (h) rt M2, (i) rt P3, (j) lt P3, (k) lt M1 Lt M2
Rt P4 frag.
Lt I1
Lt M3
Rt maxilla frag. with P3-P4
Anatomical part
Upper Laetolil Beds
Lt C1
Unknown
1935
M.42323
L.S.B. Leakey Expedition Kohl-Larsen Expedition Kohl-Larsen Expedition Kohl-Larsen Expedition M. Mwoka
Table 7.2 Catalogue of fossil hominin specimens recovered from Laetoli Specimen Year Collector Locality Stratigraphic position
A. afarensis A. afarensis Homo cf. H. sapiens
A. afarensis
A. afarensis
A. afarensis A. afarensis A. afarensis
A. afarensis A. afarensis A. afarensis
A. afarensis A. afarensis
A. afarensis
A. afarensis
A. afarensis
A. afarensis
A. afarensis
A. afarensis
A. afarensis
A. afarensis
Taxon
(continued)
Incorrectly listed as Loc. 25 in Leakey (1987a)
Incorrectly stated as 3 m above Tuff 5 in Leakey (1987a); see White (1980a: 503) Incorrectly stated as 9 m above Tuff 8 in Leakey (1987a); see White (1980a: 503). Previously identified as M3 (White 1980a) Lt M1 in Leakey (1987a)
Same individual as L.H. 22 Incorrectly stated as 5.5 m above Tuff 3 in White (1980a)
Same individual as L.H. 27 and 28
Lectotype of A. afarensis (see Table 7.1)
Found by M. Muoka in Leakey et al. (1976)
Published as M.18773 (White 1980a, 1981) Holotype of Praeanthropus (see Table 7.1)
Comments
7 Hominins from Laetoli 145
1976
1977 1978
1978 1978
1978
1979 1979 1979
1975
L.H. 21
L.H. 22 L.H. 23
L.H. 24 L.H. 25
L.H. 26
L.H. 27 L.H. 28 L.H. 29
L.H. 30
1998
M.D. Leakey Expedition L. Dotha M.D. Leakey Expedition M.D. Leakey Expedition C. Robinson
N. Mbuika P. Sila M. Mwoka
J. Masovo
E. Kandindi M. Mwoka
E. Kandindi M. Mwoka
Loc. 22S
Loc. 10 Loc. 7E (strip 8) Loc. 8
Loc. 7
Loc. 8 Loc. 8 Loc. 8
Loc. 6
Loc. 10E Loc. 2
Loc. 11 Loc. 8
Loc. 12E
Loc. 8 Loc. 1
Locality
Upper Ndolanya Beds
Unknown
Probably between Tuffs 6 and 8 Unknown Upper Ndolanya Beds
~2 m below Tuff 7 ~2 m below Tuff 7 Unknown
~3.7 m below Tuff 7
~2.1 m below Tuff 7 15 cm above Tuff 6
~0.9 m above Tuff 7 ~1.3 m below Tuff 7
Between Tuffs 6 and 7
Between Tuffs 5 and 6 Between Tuffs 7 and 8
Stratigraphic position
Taxon
Comments
Lt proximal tibia
Rt M3 Rt zygomatic process of frontal; two associated bone frags. (infant) Rt C1
A. afarensis cf. Hominini indet. A. afarensis
Previously undescribed
Specimen lost Previously identified as a cercopithecid
A. afarensis Non-hominin, Initially identified as a hominin, but subsequently recognized as a Rhinocolobus cercopithecid A. afarensis Incorrectly stated as from Loc. 12 Partial skeleton (juvenile): (a) rt maxilla, (b) lt (White 1980a) maxilla frag., (c) lt maxilla frag, (d) lt dP4, (e) rt zygomatic, (f) lt frontal frag., (g) frontal frag., (h) parietal frag., (i) rt parietal frag., (j) parietal frag., (k) rt temporal frag., (l) temporal frag., (m) rt occipital frag., (n) lt occipital frag., (o) rt occipital frag., (p) rt clavicle frag., (q) rib sternal end, (r) lt proximal ulna, (s) ulna frag., (t) distal ulna frag., (u) rt femur shaft and neck, (v) lt femur shaft frag., (w) intermediate phalanx, (x) proximal phalanx, (y) proximal phalanx, (z) metacarpal II frag., (a-1) metapodial head, (a-2) phalanx epiphysis Lt P4 and M1 A. afarensis Same individual as L.H. 8 Lt M2 A. afarensis 2 m below Tuff 7 according to the field catalogue entry Lt P3 A. afarensis Rt P3 A. afarensis From Loc. 14 according to White (1980a). Incorrectly listed as Rt P3 in Leakey (1987a) Rt M2 A. afarensis Finder is J. Masobo according to White (1980a) Rt M3 A. afarensis Same individual as L.H. 5 and 28 Rt M2 A. afarensis Same individual as L.H. 5 and 27 Lt mandibular frag. with M1-M3 A. afarensis Published as Homo cf. H. erectus (Leakey 1987a) Lt dC1 A. afarensis Synonymous with L.H. 3/6 (c)
Lt M2 Lt I1
Anatomical part
Hominini gen. et sp. indet. EP 162/00 2000 A. Kweka Loc. 16 Between Tuffs 7 and 8 Lt C1 A. afarensis EP 2400/00 2000 M. Mbago Loc. 16 51 cm above Tuff 8 Rt mandibular frag. with P3-M1 A. afarensis EP 1500/01 2001 T. Harrison Silal Artum Upper Ndolanya Beds Rt maxillary frag., edentulous P. aethiopicus Sources: White (1977, 1980a, 1981); Day et al. (1980b); Magori and Day (1983); Leakey (1987a); Kyauka and Ndessokia (1990); Kyauka (1994); Ullrich (2001); Harrison (2002, unpublished data)
EP 1000/98
LAET 79–5447 1979
L.H. 31 1987 LAET 75–3817 1975
M. Mwoka C. Kamau
1976 1976
L.H. 19 L.H. 20
M. Mwoka
Collector
Table 7.2 (continued) Specimen Year
146 T. Harrison
7 Hominins from Laetoli
147
original stratigraphic unit from which the hominin fossil was derived (usually narrowly constrained between two sequential marker tuffs, such as between Tuffs 7 and 8 in the case of EP 2400/00). However, except for those rare instances of hominins being found in situ (i.e., L.H. 2, L.H. 3 and L.H. 6), it is important to make a distinction between the finding spot and the presumed original stratigraphic placement of the specimens. For example, Leakey (1987b) recorded precise information on the stratigraphic location of the hominins recovered by her team, but this represents the stratigraphic level of the surface on which specimens were found rather than that of the level from which they eroded. Although long distance transportation can be largely discounted at Laetoli, the displacement of surface finds by livestock and game animals and by seasonal run-off over short distances is certainly conceivable, and can be shown to have occurred for some associated dental remains (Leakey 1987b). The main point is that all surface finds, including those recovered by Mary Leakey for which precise stratigraphic locations have been published, are at best derived from narrow stratigraphic horizons between marker tuffs. The three hominin specimens recovered by Kohl-Larsen are certainly derived from the Upper Laetolil Beds based on their preservation, but otherwise the published and archival information does not permit a more precise geographic or stratigraphic provenance (Kohl-Larsen 1943; Protsch 1981). Garusi Hominid I and II were apparently recovered from the same locality, 16 days apart. Protsch (1981: 12) has described the location as being “the most northwesterly corner of the Garusi River”, but Kohl-Larsen’s sketch map (published in Protsch 1981: 4) marks the find spot to the northeast of the
head of the Garusi River, between the Garusi and Gadjingero river valleys. Protsch (1981: 12) further indicates that “the finds were located about 500 m west of Kohl-Larsen’s campsite at a tributary of the Garusi River, at the foot of [a] … sandstone plateau” (see also Kohl-Larsen 1943: 386). From the sketch map published by Kohl-Larsen (1943), we know that his camp (Lagerplätze) was situated in the vicinity of Loc. 4 on the southern side of the Garusi valley. All of the evidence, which is admittedly rather scanty, appears to be consistent with the Garusi hominins having come from Loc. 16. Protsch (1981: 10–11) published photographs from the Kohl-Larsen expedition that identify the location of the Garusi I and II finds. Unfortunately, I have not been able to relocate the exact spots where these photographs were taken because there are no distinguishable landmarks, but the photos are not inconsistent with them having been taken at Loc. 16. Another issue pertaining to the provenance of the hominin fossils collected by Mary Leakey is the stratigraphic placement of the marker pedestals at Laetoli. The hominin find spots were marked by stones embedded in a concrete block with the L.H. number inscribed on top (Leakey 1987b). Most of these pedestals are still traceable today, but the original structures have been damaged to varying degrees, and in some cases they have been repaired or replaced by subsequent workers (Mabulla 2000). The problem is that the locations of the pedestals do not always match the recorded stratigraphic position of the hominin find. According to new observations in the field (see Ditchfield and Harrison 2011) there is a discrepancy in the stratigraphic location of at least seven pedestals (Table 7.3). There are two possible
Table 7.3 Discrepancies between the published stratigraphic position of hominins at Laetoli and the placement of the marker pedestals Recorded stratigraphic position of hominin Stratigraphic position of marker (Leakey 1987b) pedestal Additional comments Specimen Locality L.H. 3/6 L.H. 4 L.H. 7
Loc. 7 Loc. 7 Loc. 5
Between Tuffs 7 and 8 On top of Tuff 6 1.2 m below Tuff 7 1.25 m below Tuff 6 ~0.6 m above Tuff 5 On top of Tuff 3
L.H. 10
Loc. 10W
5.5 m below Tuff 3
Not located
L.H. 11
Loc. 10W
7.3 m below Tuff 3
Not located
L.H. 12
Loc. 5
1.8 m below Tuff 4
Just below Tuff 3
L.H. 14
Loc. 19
~0.3 m above Tuff 5
~0.3 m above Tuff 5
L.H. 15 L.H. 16 L.H. 21 L.H. 23 L.H. 25 L.H. 26
Loc. 1 Loc. 6 Loc. 12E Loc. 8 Loc. 2 Loc. 6
9 m above Tuff 8 Just below Tuff 6 Between Tuffs 6 and 7 ~1.3 m below Tuff 7 15 cm above Tuff 6 ~3.7 m below Tuff 7
~0.9 m above Tuff 8 1.5 m below Tuff 6 7.6 m above Tuff 7 Between Tuffs 6 and 7 Not located 5 m below Tuff 6
See Ditchfield and Harrison (2011) See Ditchfield and Harrison (2011) See Ditchfield and Harrison (2011). The main fossilbearing horizon at Loc. 5 is between Tuffs 3 and 5 Incorrectly stated as 5.5 m above Tuff 3 in White (1980a). Based on the section in Hay (1987) the hominin is from between Tuffs 1 and 2. This is the main fossil-bearing horizon at Loc. 10W Based on the section in Hay (1987) the hominin is from between Tuffs 1 and 2. This is the main fossil-bearing horizon at Loc. 10W The main fossil-bearing horizon at Loc. 5 is between Tuffs 3 and 5 Incorrectly stated as 3 m above Tuff 5 in Leakey (1987a); see White (1980a: 503) Typographic error in Leakey (1987a); see White (1980a: 503) Incorrectly stated as from Loc. 12 (White 1980a) 2 m below Tuff 7 according to the catalogue entry From Loc. 14 according to White (1980a) See Ditchfield and Harrison (2011)
148
T. Harrison
explanations for these inconsistencies: (1) the stratigraphic placement of the hominin is incorrectly recorded; or (2) the pedestals were placed in the wrong positions. Without evidence to the contrary, I am inclined to accept that the recorded position of the hominins is accurate, and that the pedestals are incorrectly placed. Given these considerations, as well as what is known about the occurrence of fossiliferous horizons in the Upper Laetolil Beds (see Harrison and Kweka 2011), most of the fossil hominins can be placed into their appropriate statigraphic context (see Table 7.4). It can be seen that A. afarensis specimens occur throughout the Upper Laetolil Beds, with dates ranging from 3.63 Ma to 3.83 Ma. When their stratigraphic placement is taken into consideration, it can be seen that most of the hominins from the Upper Laetolil Beds (n = 23; 88.5%) are derived from above Tuff 5, and there are relatively few specimens from the lower part of the sequence. However, this is largely a reflection of the number of exposures and the frequency of occurrence of fossil vertebrates throughout the sequence. The percentage of fossil mammals recovered from above Tuff 5, between Tuffs 3 and 5, and below Tuff 3 are 80.6%, 5.5% and 13.9% respectively. By comparison the corresponding percentages of hominin finds are 88.5%, 3.9% and 7.7%, which implies that their frequency of occurrence is quite close to that expected given variations in paleontological productivity throughout the sequence. However, the relative rarity of hominins obtained from the productive localities that sample the sequence below Tuff 3 (i.e., Locs. 10, 10W and 9S) may eventually prove to be of some significance.
Table 7.4 Stratigraphic distribution of hominins in the Upper Laetolil Beds Marker Tuff Age (Ma)a In situ findsb Surface findsb Yellow Marker Tuff 3.63 L.H. 15, EP 2400/00 Tuff 8
Since 1998, two additional specimens of A. afarensis have been recovered, both derived from the upper part of the Upper Laetolil Beds at Loc. 16: EP 2400/00, a right mandibular fragment with P3-M1, and EP 162/00, a left lower canine. In addition, two specimens recovered by Mary Leakey’s expeditions are described here for the first time: L.H. 29, a left mandibular fragment with M1-M3, and LAET 79-5447, an isolated upper canine. An isolated lower incisor, MB Ma. 8294 (Garusi 4), in the Kohl-Larsen collections of the Museum für Naturkunde in Berlin, briefly discussed by Ullrich (2001), is also formally described here for the first time.
EP 2400/00 Right mandibular fragment with P3-M1 (Fig. 7.3). Location and Stratigraphic Provenance This specimen was discovered by Michael Mbago at Loc. 16 on August 7, 2000. The specimen was found on the surface in a shallow drainage channel on the western flank of the main gully, 51 cm above Tuff 8 (see Harrison and Kweka 2011). Although EP 2400/00 had been displaced from it original location at the time of discovery, its original stratigraphic provenance can be interpreted to be 0.5–1.3 m above Tuff 8. The absolute age of specimen can be constrained by the bracketing radiometric dates for Tuff 8 (3.631 ± 0.018 Ma) and the overlying Yellow Marker Tuff (3.627 ± 0.018 Ma) (Deino 2011), and can inferred to have an age of ~3.63 Ma (see Fig. 7.1). It is one of the youngest specimens of A. afarensis known from Laetoli.
3.63 L.H. 2, 3/6
Tuff 7
New Hominins from the Upper Laetolil Beds
L.H. 1, 8, 22, EP 162/00
3.66 L.H. 4, 5, 13, 21, 23, 24, 25, 26, 27, 28
Tuff 6
3.70
Tuff 5 Tuff 4
3.79 3.79
Tuff 3 Tuff 2
3.80 3.81
L.H. 7, 14, 16, 19
L.H. 12
L.H. 10, 11 Tuff 1 3.83 Lower Laetolil Beds 3.85–4.36 a Data from Deino (2011) b Data from Leakey (1987a) and Harrison (unpublished)
Preservation The mandibular corpus is partially preserved below the cheek teeth; the inferior portion of the corpus is missing below the level of the mental foramen. Anteriorly, the posterior margin of the alveolus for the lower canine is preserved. Posteriorly there is a pair of indentations that represents the anterior margin of the alveolus for the mesial root of M2. Laterally, the bone is quite weathered and the surface is marked by numerous fine cracks oriented anteroposteriorly in line with the bone fibers. Thin flakes of bone have been lost from the surface immediately over the mesial roots of P3 and P4, and to a lesser extent on the M1 anterior root. As a consequence, the roots are exposed to a greater degree than they would
7 Hominins from Laetoli
149
mandible was fragmented prior to fossilization, although the sharp breaks inferiorly and anteriorly suggest that additional breakage occurred after it was fossilized and exposed on the surface. There are no indications of carnivore or rodent gnawing. The teeth are very worn occlusally. P3 has lost a small flake of enamel from its distolingual margin. P4 has lost the enamel from its entire buccal face, and the marginal enamel is chipped and flaked. No enamel remains on the occlusal surface of M1. In addition, enamel has been chipped from the mesiobuccal corner of the crown, as well as from the mesial and lingual faces. However, preservation is adequate to allow accurate measurements of the original dimensions of the teeth.
Morphology
Fig. 7.3 EP 2400/00, right mandibular fragment with P3-M1 of Australopithecus afarensis. (a) lateral view; (b) occlusal view; (c) medial view
have in life. Posteriorly, a large triangular flake of bone has been removed by abrasion from the lateral side of the corpus from just below M1, at the point where the corpus is beginning to expand for the anterior root of the ramus. The medial side of the corpus also shows fine longitudinal cracking of the surface bone, with a relatively sharp break anteriorly, and a feathered surface posteriorly below and behind M1, which may be the result of pre-fossilization weathering. None of the broken edges are fresh, and it seems likely that the
The mandibular fragment consists of the alveolar portion of the corpus below P3-M1. The inferior portion of the corpus is missing, with only a maximum depth of 18.6 mm remaining below P4-M1. The corpus appears to be relatively thick in relation to the size of the teeth. The thickness of the alveolar portion of the corpus is 18.3 mm at mid-P3, 18.9 mm at mid-P4, and at least 22.2 mm at mid-M1. The lateral surface of the corpus below P3 is relatively flat. A small accessory mental foramen, with a diameter of 1.7 mm, is located vertically 15.2 mm below mid-P3. There is a shallow groove posterior to the aperture indicating that a branch of the mental nerve exited in this direction. The corpus is broken at the level of the main mental foramen, but its superior and posterior margins of the mental foramen are preserved. It was quite large and elliptical in shape, with a minimum anteroposterior diameter of 4.8 mm. The preserved margins are sharply defined. The dorsal margin of the foramen is located 19.4 mm below the cemento-enamel junction (and 17.5 mm below the alveolar margin) of the mesial margin of P4. The accessory foramen is situated superior and anterior to the main foramen, and separated from it by a distance of 6.3 mm. Directly superior to and slightly posterior to the main foramen is a distinct, but shallow depression, which occurs just inferior to the root apices of P4. Superior to this the lateral aspect of the corpus is slightly convex inferosuperiorly. Below M1 the lateral wall is more markedly convex, where the lateral prominence of the anterior root of the ramus originates. In inferior view, the broken surface shows that the bony wall of the corpus was thick, up to 3.6 mm thick both medially and laterally. Medially the surface of the corpus is evenly convex infero-superiorly; more markedly so below P3 than M1. Posteriorly the alveolus for the mesial root of M2 indicates that it was a relatively broad tooth buccolingually. The minimum breadth of the alveolus is 13.7 mm, which compares with the equivalent dimension in M1 of only 12.1 mm. The alveolus for the
150
lower canine is partially preserved. It has a minimum diameter of 8.0 mm, but judging from its contour it would have exceeded 9 mm in diameter. It was clearly a sizeable root, although it did not exceed the span of the roots for P3. The inferosuperior length of the canine alveolus is incomplete, but at its point of breakage, 12.3 mm below the superior margin of the alveolus, it still accommodated a stout root. Judging from the relatively small size of the alveolus for the canine root EP 2400/00 likely belonged to a female individual. The minimum distance between the alveoli of the canine and the mesial root of P3 is only 1.6 mm, so there was effectively no diastema. In occlusal view, the preserved cheek tooth row exhibits a very slight lateral convexity. Judging from the preserved portion of the canine alveolus, the canine crown would have been positioned slightly medial to the long-axis of the postcanine tooth row, as is typical of A. afarensis compared to the more primitive lateral position in A. anamensis (Ward et al. 2001). The P3 is larger in overall size than P4. It is obliquely oriented, with its long-axis oriented at 44° to the line of P3-M1. The maximum length of the crown is 11.9 mm and its perpendicular breadth is 9.0 mm. The mesiodistal and buccolingual dimensions are given in Table 7.6. In occlusal view, the crown is trapezoidal in shape, with the greatest breadth mesially and narrower distally. The occlusal surface of the crown is worn nearly flat, with much of the enamel worn away and an extensive area of dentine exposed (more than 50% of the occlusal surface). Apart from the enamel around the margin of the crown, there is a small area retained at the base of the crown mesiobuccally and a small circular area (3.5 × 3.2 mm) of very thin enamel retained in the talonid basin. The protoconid is worn flat, and is represented by a large area of exposed dentine, but was almost centrally positioned on the crown. The mesiobuccal face of the crown bulges outwards beyond the lateral surface of the corpus, but then tapers apically. The inferior enamel junction on the mesiobuccal face extends slightly more inferiorly onto the root than it does in the rest of the crown. The worn height of the mesiobuccal face is 6.1 mm, compared with only 3.3 mm lingually. The small size and morphology of the P3 confirms the observation from the size of the alveolus for the canine root that EP 2400/00 belongs to a female individual. Occlusal contact between the upper canine and the mesiobuccal face of P3 was evidently concentrated on the apex of the protoconid, and did not extend far down onto the mesiobuccal face of the crown. An obliquely oriented furrow in the enamel at the base of the crown mesially represents a trace of the buccal cingulum. A vestige of cingulum is also present along the buccal margin of the crown, and these two sections are linked by a shallow and irregular indentation in the enamel surface that arcs mesiodistally around the mesiobuccal face, about 4 mm up from the base of the crown. The buccal face of the crown is evenly convex, except for a slight angulation midway along its length, presumably representing the point where the postprotocristid would have converged with the distal marginal ridge. Mesiolingually, the
T. Harrison
crown has a prominent protuberant beak, which represents the mesial termination of the preprotocristid. The mesiolingual margin of the crown is slightly concave. Distolingually there is a small triangular area of dentine exposure, continuous with that produced by the worn protoconid, located between the island of enamel preserved in the talonid basin and a small fold of enamel along the mesial margin that represents a remnant of enamel between the preprotocristid and hypoprotocristid. This area of dentine exposure represents the location of the metaconid, which, in the unworn state, would have been very small judging from the size of the dentine exposure. The talonid heel was narrow. Distally, where the enamel wall is exposed, the enamel is 1.1 mm thick. The tooth is two-rooted, with the mesial and distal roots subequal in size. The mesial root is placed more laterally than the distal root, and it is cylindrical in shape rather than mesiodistally compressed. The exposed portion of the mesial root has a length of 10.2 mm, but it cannot have been longer than 15 mm, otherwise it would be visible inferiorly where the corpus is broken. The P4 is oval in occlusal outline, with a long-axis directed at 72° to the line of P3-M1. The crown is broader than long. The tooth is heavily worn, with dentine exposure over more than 50% of the occlusal surface. The protoconid and metaconid have been worn flat and are represented by contiguous areas of dentine exposure. The two cusps were apparently subequal in size. Two thin layers of enamel are retained on the flattened occlusal surface. A large rectangular remnant is located centrally and distally, corresponding to the floor of the talonid basin, and a much smaller oval-shaped remnant occurs along the mesiolingual margin of the crown, corresponding to the floor of the mesial fovea. The mesiobuccal face of the crown bulges laterally, and its cemento-enamel junction extends inferiorly below that of P3 and M1. Even in its very worn state, the buccal face of the crown is much higher than the lingual face (7.4 mm as opposed to 2.4 mm). The tooth has two subequal roots, both transversely aligned. M1 is very heavily worn, with no enamel remaining on the occlusal surface. The specimen evidently belonged to an aged individual. The dentine surface is smoothly concave, with no residual topography of the cusps, surrounded by an elevated rim of enamel. The broken enamel suggests that the sides of the teeth were coated with relatively thin enamel, although no measurements can be taken. The crown is relatively broad and rectangular in shape, with a slight degree of buccolingual waisting midway along its length (slightly more pronounced on the buccal side). There are no observable traces of cingulum, but if originally present they likely would have been removed by the excessive wear.
Comparisons The corpus of EP 2400/00 is similar in contour and size to that of L.H. 4, and of the larger mandibles from Hadar, such as A.L.
7 Hominins from Laetoli
400-1 and A.L. 266-1, and more robust than the smaller Hadar mandibles, such as A.L. 128-23 and A.L. 198-1. The mean mediolateral breadth of the corpus below M1 in the Hadar sample (n = 22) is 19.8 mm, with a range of 15.8 mm to 24.7 mm (Lockwood et al. 2000; Kimbel et al. 2004) (Table 7.5). EP 2400/00 with a minimum breadth at M1 of 22.2 mm places the Laetoli specimen in the upper end of the range for A. afarensis and A. anamensis, and close to the mean value for A. africanus (Tobias 1991; Ward et al. 2001; Kimbel et al. 2004). Judging from the contour of the broken anterior margin of the medial surface of the corpus, the mandible would have had an anteroposteriorly elongated subincisive planum, as in other specimens of A. afarensis, such as A.L. 400-1 and A.L. 198-1, and more pronounced than in A.L. 266-1 and A.L. 288-1. However, it was clearly not as posteriorly inclined as in A. anamensis (Ward et al. 2001). The medial surface of the corpus below M1 is strongly convex, indicating that it was heavily buttressed medially as in A.L. 400-1. The root of the ramus on the lateral side of the corpus is situated below M1, as in L.H. 4, MAK-VP-1/12,
151
and most of the mandibular specimens from Hadar (White et al. 2000; Kimbel et al. 2004), but it does occur more posteriorly in A.L. 400-1 (mesial M2) and A.L. 198-1 (distal M2). EP 2400/00 has a small elliptical depression located below mid-P4, just superior and posterior to the mental foramen. A similar small depression occurs in A.L. 207-13, A.L. 288-1i, A.L. 333w-60, A.L. 400-1a, A.L. 437-2, A.L. 438-1 g, A.L. 444-2, L.H. 4 and MAK-VP-1/12, and the general area is concave in A.L. 198-1, but otherwise mandibles of A. afarensis are uniformly convex in this area. The mental foramen is positioned below mesial P4. The modal position in A. afarensis and A. africanus is below P4, although it varies in location from below distal P3 to P4/M1 (Ward et al. 1982; Tobias 1991). A similar pattern characterizes the small sample of mandibular specimens of A. anamensis (Ward et al. 2001). The occurrence of a main foramen and a smaller accessory foramen in EP 2400/00 is commonly observed among A. afarensis. Robinson (2003) recorded multiple mental foramina in 36.0% of the specimens from Hadar. Paired foramina also
Table 7.5 Comparison of dimensions (mm) of the mandibular corpus in EP 2400/00 with other specimens of A. afarensis from Laetoli, Hadar and Maka Height mental foramen to alveolar margin Locality Specimen Breadth at P4 Breadth at M1 Laetoli
EP 2400/00 18.9 22.2(–) 17.5 L.H. 4 18.5 19.7 21.4 Hadar A.L. 128-23 16.6 18.0 17.4 A.L. 145-35 18.9 21.1 18.0 A.L. 198-1 15.8 15.8 18.4 A.L. 207-13 17.4 18.1 21.0 A.L. 228-2 16.0 16.3 20.1 A.L. 266-1 19.9 21.7 18.4 A.L. 277-1 17.8 17.9 23.0 A.L. 288-1i 16.6 17.1 20.0 A.L. 311-1 22.0 – 26.3 A.L. 315-22 17.3 19.2 21.1 A.L. 330-5 18.5 20.9 19.7 A.L. 333w-12 16.8 17.4 19.0 A.L. 333w-1a,b 18.9 19.4 18.7 A.L. 333w-32 + 60 22.0 23.6 24.1 A.L. 400-1a 18.5 18.7 20.1 A.L. 417-1a 18.4 18.0 21.5 A.L. 433-1a 20.3 20.2 17.0 A.L. 437-1 21.2 20.0 25.0 A.L. 437-2 22.2 22.2 23.2 A.L. 438-1 g 25.0 24.7 20.5 A.L. 444-2 21.1 23.0 21.6 A.L. 582-1 22.6 21.4 21.1 A.L. 620-1 19.5 20.5 23.5 Maka MAK-VP 1/12 17.7 18.7 18.8 Dimensions: Maximum mediolateral breadth of corpus at mid-P4; maximum mediolateral breadth of corpus at mid-M1; vertical inferosuperior height of the corpus between the alveolar margin and the mental foramen Data: Laetoli (Harrison, unpublished); Maka (White et al. 2000); Hadar (Kimbel et al. 2004). Where both sides are measurable, the value is the average of the right and left sides.
152
occur in the MAK-VP-1/12 mandible from Maka and in L.H. 4 from Laetoli. In EP 2400/00 the mental foramen is located 17.5 mm below the alveolar margin (Table 7.5). This falls at the low end of the range for the A. afarensis sample from Hadar (17.0–26.3 mm, n = 23), which has a mean value of 20.8 mm (Kimbel et al. 2004). The distance in L.H. 4 is 20.4 mm and 22.4 mm on the left and right sides of the corpus respectively. Although incomplete, the canine alveolus in EP 2400/00 is consistent in size with the canine root in EP 162/00, which has a mesiodistal length of 9.3 mm and breadth of 6.3 mm. The small size of the alveolus for the canine root would suggest that EP 2400/00 belonged to a female individual. This is supported by the size and morphology of P3. P3 in EP 2400/00 is the smallest known example of this tooth from Laetoli (Table 7.6). It is slightly smaller than that in L.H. 4 and L.H. 2, which have occlusal areas 1.6% and 2.0% larger respectively, but it is much smaller (18.8% smaller in area) than the very large P3 in L.H. 3. The morphology of the crown in EP 2400/00 is quite similar to that in L.H. 4 and L.H. 24, but it does differ in a number of respects: more pronounced mesiolingual beak, narrower distal basin, smaller metaconid, more pronounced buccal cingulum, and greater extension of the cemento-enamel junction mesiobuccally onto the base of the mesial root. EP 2400/00 is similar in occlusal outline to L.H. 3, but it is much smaller, and has a more prominent mesiolingual beak, a less convex distal margin, and probably a much smaller metaconid. However, the latter is similar in having distinct traces of the buccal cingulum mesially and distally. Also, the P3 in L.H. 4 is oriented less obliquely to the long axis of the cheek tooth row (28°) than in EP 2400/00 (44°), which is close to the mean value (43°, range = 32–52°) for the Hadar sample. Judging from the orientation of the distal contact facet in L.H. 24, the crown was positioned more obliquely in the tooth row than in EP 2400/00. The configuration of the roots appears to match that in L.H. 4 (and the majority of specimens from Hadar), with a large cylindrical mesial root and a mesiodistally compressed distal root. However, as noted by White et al. (2000) and Kimbel and Delezene (2009) there is variation in P3 root number and structure at Laetoli and Hadar, ranging from a pair of roots as in EP 2400/00, to a Tome’s root (e.g., L.H. 14, A.L. 145-35, A.L. 288-1, A.L. 400-1a), and divided distal root (e.g., L.H. 24) The P3 in EP 2400/00 is quite similar in size and shape to the smaller examples of P3 from Hadar, such as A.L. 128-23 and A.L. 288-1, which presumably belonged to female individuals. It differs from A.L. 288-1 in being slightly larger in size, mesiodistally longer, and with a more distinct lingual cingulum, especially mesially. EP 2400/00 is similar in shape, proportions and general morphology to A.L. 128-23, but it is slightly larger in size. The specimens are a good match in the development of the mesiobuccal beak and the apparent small size of the
T. Harrison
metaconid. A.L. 128-23 differs in having a more protuberant distal tubercle and a less well-developed buccal cingulum mesially. EP 2400/00 is also larger than the P3 in A.L. 207-13, but similar in proportions. It differs in having a more pronounced mesiolingual beak, a less protuberant distal tubercle, a more distinct buccal cingulum mesially, and probably a smaller metaconid. A.L. 400-1 is similar in overall size to EP 2400/00, but it had a larger metaconid, a less-well-developed buccal cingulum, and it lacks the mesiolingual beak. A.L. 277-1 is slightly larger, with a smaller mesiolingual beak, a somewhat larger metaconid, and a less shelf-like buccal cingulum. EP 2400/00 is similar in size to A.L. 266-1, but it has a narrower and longer crown, a more prominent mesiolingual beak, a less protuberant distal tubercle, and a slightly smaller metaconid. As noted by White (1985), the full range of metaconid expression, from absent to well developed, is present in the sample from Hadar. A weak or absent metaconid is found in 40.0% of P3s from Hadar. In the previously collected hominins from Laetoli, a large metaconid is present in L.H. 2, L.H. 3, L.H. 4, and L.H. 14, whereas it is weakly developed in L.H. 24 (20.0% of the sample). EP 2400/00 adds a second example of P3 from Laetoli with a weakly expressed metaconid, bringing the incidence to 33.3%. It may be that a lower proportion of P3s with a well-developed metaconid does occur at Hadar compared with the specimens from Laetoli, but the samples are still too small to adequately test the significance of the difference. Overall, the P3 of EP 2400/00 does not appear to have any morphological features that can be used to consistently discriminate it from the sample from Hadar, except that the crown is relatively longer. Lockwood et al. (2000) and Kimbel et al. (2006) have demonstrated that the length of P3 in the Laetoli sample is significantly greater than that in the Hadar sample, and that this is part of a temporal trend in A. afarensis. The more primitive condition, in which the P3s are relatively longer than in A. afarensis from Laetoli, is seen in Australopithecus anamensis (Ward et al. 2001). With the recovery of EP 2400/00, a relatively small P3, which is more similar in overall size to examples from Hadar, the magnitude of the temporal trend is somewhat diminished. Nevertheless, the mesiodistal length and the maximum length of the P3 crown are still significantly greater in the Laetoli sample than in the sample from Hadar (see Discussion). The P4 in EP 2400/00 is consistent in length and breadth dimensions to previously described specimens of A. afarensis (Table 7.6). In terms of its occlusal area (mesiodistal length ´ buccolingual breadth; 103.7 mm2) EP 2400/00 falls in the lower end of the range for the Hadar sample (mean = 106.9 mm2; range = 77.0–134.5 mm2; Kimbel et al. 2004), being most comparable in size to A.L. 228-2, A.L. 266-1 and A.L. 4001a. EP 2400/00 is also smaller than the P4 in L.H. 3 and L.H. 14. It is similar in size those of L.H. 4, but the crown is slightly shorter. The long-axis of the P4 in EP 2400/00 relative
7 Hominins from Laetoli
153
Table 7.6 Dimensions (mm) of EP 162/00 and EP 2400/01 compared with other A. afarensis teeth from Laetoli, Hadar, Dikika and Maka Laetolib Hadar and Dikikac Makad a Tooth Dimension EP N Mean Range N Mean Range N Mean Range C1
MD 8.5 2 10.5 9.3–11.7 10 8.6 7.5–9.5 1 9.5 9.5 BL 8.0 3 10.3 10.1–10.4 11 10.6 8.8–12.4 1 10.2 10.2 P3 MD 10.8 6 10.9 9.8–12.2 19 9.2 7.9–11.4 2 9.5 9.3–9.7 BL 10.0 6 10.9 9.8–12.3 19 10.4 8.9–12.6 2 11.3 11.2–11.3 P4 MD 9.1 5 10.4 9.6–11.1 20 9.7 7.7–11.4 2 9.4 9.0–9.7 BL 11.4 5 11.5 10.8–12.1 17 11.0 9.8–12.8 2 10.4 9.9–10.8 M1 MD 12.3 5 13.4 12.2–14.2 23 13.0 10.1–14.8 3 13.2 13.0–13.6 BL 11.4 5 13.1 12.5–13.5 17 12.5 11.0–13.5 3 12.2 12.1–12.4 MD mesiodistal length, BL buccolingual breadth a New specimens collected from Laetoli by the Eyasi Plateau expedition (1998–2005): EP 162/00 (lower canine) and EP 2400/00 (mandible with P3-M1). Canine measurements follow method used by White (1977), but discrepancies reflect differences in tooth orientation b Specimens from Laetoli collected by Louis Leakey (1935) and Mary Leakey (1974–1979). Canine data from White (1977, 1980a); all other measurements by the author c Data from Kimbel et al. (2004) and Alemseged et al. (2006) d Data from White et al. (2000). The data include estimated and maximum values, as well as teeth from both sides of the MAK-VP-1/12 mandible
to the line of the cheek teeth (72°) is more obliquely directed than the majority of P4s from Hadar, which have a mean orientation of 61°, but it does fall in the upper end of the range (43°–85°). At noted by White (1985), and confirmed by further comparisons of EP 2400/00, there appear to be no consistent differences in the morphology of P4 in the Hadar and Laetoli samples. M1 in EP 2400/00 is heavily worn and prevents detailed comparison of the occlusal morphology. In terms of its mesiodistal and buccolingual dimensions it represents the smallest example from Laetoli, and falls in the low end of the range of the series of first lower molars from Hadar (Table 7.6). The crown is relatively narrow, with a breadthlength index of 92.7, which again falls at the low end of the range for the sample from Hadar (mean = 95.8, range = 88.5– 103.1; Kimbel et al. 2004).
EP 162/00 Left lower canine (Fig. 7.4).
Location and Stratigraphic Provenance This specimen was found by Amandus Kweka at Loc. 16 on January 17, 2000. It was recovered as a surface find between Tuffs 7 and 8 (see Harrison and Kweka 2011). The absolute age of EP 162/00 can be constrained by the new radiometric dates for Tuff 7 (3.67 ± 0.04 Ma) and Tuff 8 (3.631 ± 0.018 Ma) (Deino 2011), giving the specimen an inferred age of ~3.63– 3.67 Ma (see Fig. 7.1).
Fig. 7.4 EP 162/00, left lower canine of Australopithecus afarensis. (a) buccal view; (b) lingual view
Preservation The crown is damaged by abrasion and slightly worn. A section of enamel 3.7 mm wide has been lost from the
154
distal margin around the base of the crown, so that the distal tubercle is incompletely preserved. In addition, a large flake of enamel has been lost from the mesial and mesiobuccal face of the crown from the tip of the apex to half way down the crown. The lingual face is moderately pitted by weathering, and an apico-basally directed crack originates at the base of the crown distolingually and continues for much of the length of the root. The mesial and buccal faces of the crown are lightly pitted by weathering, with a series of fine cracks running apico-basally around the base of the crown. Mesiolingually there is a prominent crack that extends almost to the apex of the crown and runs almost the full length of the root. The root is complete, but the distal margin of the apex shows clear signs of having been gnawed by a small rodent. It is evident from the weathering and rodent gnawing that the isolated canine was exposed on the Pliocene land surface prior to being buried. The loss of enamel from the crown appears to be relatively fresh, and was likely caused by the trampling of livestock.
T. Harrison
from the apex to the mesiolingual margin of the crown. The exposed enamel has a maximum thickness of 0.7 mm. The buccal surface is mesiodistally convex and curves distolingually towards the apex. It is generally smooth and featureless, except for a shallow groove around the base of the mesiobuccal face, representing a vestige of the buccal cingulum. In addition, skirting the base of the buccal face of the crown from the mesial margin to the distal tubercle is a hypoplasia, represented by a distinct band of thin enamel, 0.8 mm wide and originating about 2.4 mm up from the base of the crown. The root is apico-basally much taller than the crown (the length of the root buccally is 22.1), and relatively stout. It is elliptical in cross-section, with a slightly concave mesiolingual face. In lateral view the distal margin is relatively straight, while the mesial margin is convex and curves distally. In distal view the root is relatively straight, although the apex shows a slight curvature towards the lingual side.
Comparisons Morphology The crown is relatively short and distally recurved. The height of the crown is estimated to be 11.7 mm (with the chipped apex it has a minimum height of 11.4 mm). The mesiodistal length and buccolingual breadth of the crown are given in Table 7.6, along with comparative data on other specimens of A. afarensis from Laetoli and Hadar. It is oval in occlusal outline, and markedly buccolingually compressed. The breadth-length index is estimated to have been 80.9. The apex is situated distal to the midline axis of the root in the mesiodistal plane. The lingual face of the crown is apico-basally slightly concave and mesiodistally convex. It is bordered basally by a low, rounded and ill-defined lingual cingulum, which is best developed mesially. The distal margin, and most of the distal tubercle on the distobuccal margin have been lost through abrasion, but it is evident that the latter was quite prominent. A low rounded distal crest descends from the apex to terminate at the distal tubercle. Just lingual to the distal crest, and slightly better developed, is a low distolingual crest that descends from the apex to the base of the crown. The two crests are separated by a shallow crescent-shaped groove, which is deepest and broadest basally. Both crests have been flattened by slight wear along their apical aspects. Mesially the lingual cingulum curves apically. The enamel at the mesial junction of the lingual cingulum is damaged, but presumably it would have become continuous mesially with the relatively short mesial ridge. The mesial ridge itself is not preserved. The enamel in this region has been sheared away to expose a narrow strip of dentine running obliquely across the crown
EP 162/00 is comparable in size and morphology to the smaller lower canines from Hadar, which presumably belonged to female individuals (Table 7.6). EP 162/00 is most similar to A.L. 198-1, but the latter is slightly larger (although the crown height is comparable), and has a less strongly bilaterally compressed crown and a thicker and more rounded lingual pillar. The specimen also matches quite well with the incomplete crown in A.L. 128-23, which is among the smallest lower canines of A. afarensis. EP 162/00 is slightly larger and has a somewhat broader distal face, but is otherwise similar in proportions and overall dimensions. They also share a similarly placed hypoplastic feature on the buccal aspect of the crown. The larger canines from Hadar, such as A.L. 333w-58, A.L. 333-90 and A.L. 277-1, presumably from male individuals, differ in being relatively higher crowned and less bilaterally compressed, with more profound mesial and distal grooves, a more strongly developed and rounded lingual pillar, a betterdefined lingual cingulum mesially, and a more prominent distal tubercle. EP 162/00 is much smaller and relatively lower-crowned than the five examples of lower canines previously recovered from Laetoli (L.H. 2, L.H. 3n, L.H. 14e, L.H. 14f and M.42323) (Table 7.6). Compared with the range of size variation and morphology seen in the larger sample from Hadar, the latter canines from Laetoli are all likely to have belonged to male individuals. EP 162/00 represents, therefore, the first canine from Laetoli that can be assigned to a female individual, and shows that the size range of A. afarensis canines from Laetoli is comparable to that at Hadar.
7 Hominins from Laetoli
155
Morphology
Fig. 7.5 MB Ma. 8294, left I1 of Australopithecus afarensis. From left to right: lingual view; buccal view; distal view; mesial view
MB Ma. 8294 Unerupted left I1 (Fig. 7.5). Location and Stratigraphic Provenance The specimen was first identified as that of a hominin in the late 1970s by Eric Delson, while studying the Laetoli cercopithecids from the Kohl-Larsen collection in Berlin (White 1981). The root bears an original field abbreviation “gar.”, which signifies that it came from the Garusi Valley (= Laetoli). Preservation of the specimen is consistent with it being derived from the Upper Laetolil Beds, but otherwise its geographical and stratigraphic provenance is unknown.
Preservation The crown is well preserved, with no indication of wear. The enamel surface is slightly pitted due to weathering, and there are a number of fine cracks, some of which extend onto the base of the root. There are no mesial or distal interproximal contact facets. It is evident that the tooth was unerupted. The tip of the root is missing, with loss of small flakes from the distal apect of the root apex, but the mesial aspect of the root is relatively complete. Judging from the preserved portion of the root and the size of the root canal, the root was still open, but the tooth was otherwise close to being completely formed. Fine striations on the buccal, mesiobuccal and distobuccal faces of the root appear to have been produced by gnawing by a small rodent. This indicates that the tooth was isolated and exposed on the Pliocene landsurface prior to burial and fossilization.
The tooth has been briefly described by Ullrich (2001), who refers to the specimen as Garusi 4, although use of the museum accession number is preferable. The specimen is described as a lateral incisor, but the narrowness and symmetry of the crown makes it a better match with the lower central incisors from Hadar. As noted above, the specimen is well preserved and unworn. The crown is tall and relatively narrow (see Table 7.7 for dimensions). The apical margin is mesiodistally narrow, with a fine incisive edge that becomes buccolingually slightly thicker where it meets the mesial and distal margins. There are two small swellings on the distolingual aspect of the apical margin, which presumably represent traces of mammelons. The crown is mesiolingually broadest at the apex and narrows basally. The mesial and distal margins are both slightly convex, producing a relatively symmetrical crown bilaterally. The mesial margin is slightly longer than the distal margin, so the incisive apex slopes inferiorly as it passes distally. The lingual face is apicobasally strongly concave, but slightly convex mesiodistally. A low and indistinct swelling, a trace of the lingual pillar, descends the lingual face obliquely from just below the apex in the midline to blend in with the general surface of the lingual aspect of the crown about two-thirds down and to the distal side of the midline. The lingual pillar is separated from the rounded mesial and distal marginal rims by shallow grooves. There is no lingual cingulum around the base of the lingual face of the crown. In mesial view, the crown tapers apically with a concave lingual face and a slightly convex buccal face. The base of the crown meets the root at an inverted V-shaped cementoenamel junction. The root is lingually recurved, such that when the incisive apex is oriented as in tip-to-tip occlusion with its upper counterpart, the apicobasal long axis of the root would have been directed lingually at an angle of 21°. This angling of the root reflects the posterior inclination of the mandibular symphysis in A. afarensis. The buccal face of the crown is generally featureless. It is biconvex apico-basally and mesiodistally. The mesial and distal margins taper slightly in the apical half of the crown, then more abruptly in the basal half. This gives the mesial and distal margins their convexity. Overall, the crown is Table 7.7 Dimensions (mm) of I1 of Australopithecus afarensis Laetolia Hadar and Dikikab Dimensions MB Ma. 8294 N N Mean Range MD 6.3 1 7.8 6 6.3 5.6–7.1 BL 7.1 1 7.8 5 7.3 6.9–7.6 BHT 11.7(e) 1 13.0 MD mesiodistal length, BL buccolingual breadth, BHT buccal height of crown, (e) estimated value a Data collected by author b Data from Kimbel et al. (2004) and Alemseged et al. (2006)
156
T. Harrison
bilaterally almost symmetrical, apart from the slightly shorter distal margin and the slightly distally receding apex. There are no evident anomalies in the surface of the enamel associated with hypoplasias or other developmental disturbances. The root is mostly complete, except for the tip, which has lost some flakes from the mesial face. The root canal was apparently still open, with a dumbbell-shaped lumen, 4.3 mm wide in the buccolingual plane. The root is mesiodistally compressed, with shallow grooves on the mesial and distal faces. The buccal margin is slightly broader than the lingual face. The length of the root (11.4 mm) is subequal to the height of the crown, but when complete would have been slightly longer.
Comparisons MB Ma. 8294 is morphologically similar to the I1 germ in L.H. 2, the only other complete lower central incisor from Laetoli, but it is much smaller in its overall dimensions (16.9% smaller on average in its linear dimensions) (see Table 7.7). Moreover, L.H. 2 differs in being relatively broader, having a distinct median groove on the lingual face, and a greater number of mammelons on the incisive apex. L.H. 3(m) consists of the distal half of I1 only. It appears to be similar in size and morphology to L.H. 2, but slightly higher crowned, and somewhat larger in overall size than MB Ma. 8294. The I1s from Hadar are consistent in morphology and dimensions to MB Ma. 8298 (see Table 7.7) (Johanson et al. 1982; Kimbel et al. 2004).
L.H. 29 Left mandibular corpus with M1-M3 (Fig. 7.6). Location and Stratigraphic Provenance The specimen was found on July 21, 1979 by Mwongela Mwoka at Loc. 8 (field number LAET 79-5487). According to Leakey (1987b) the heavily weathered and rolled specimen was recovered from the surface of the lower part of the exposures near the Garusi River. It was assumed at the time to be derived from the Pleistocene Lower Ngaloba Beds (Leakey 1987b: 108), because of the dark staining and patination. As a consequence, the specimen was attributed to Homo cf. H. erectus (Leakey 1987b: Table 5.1). However, the specimen is heavily mineralized, unlike the Pleistocene fossils from Laetoli, which tend to be more lightly mineralized. It is much more likely that the specimen was originally derived from the fossil-rich Upper Laetolil Beds, and that it
Fig. 7.6 L.H. 29, left mandibular corpus with M1-M3 of Australopithecus afarensis. (a) occlusal view; (b) medial view
developed a similar coloration to that typical of Pleistocene fossils at Laetoli after eroding out onto the surface and becoming secondarily associated with the superficial sediments. Moreover, the morphology of L.H. 29 is entirely consistent with that of A. afarensis, and it seems most plausible to assume that the specimen was originally derived from the Upper Laetolil Beds. Regardless of its provenance, the morphological evidence alone supports reassignment of L.H. 29 to Australopithecus afarensis.
Preservation The specimen preserves the mandibular corpus below the lower molars, but it is quite heavily weathered and smoothed by rolling, especially at the broken edges. A deep longitudinal crack in the corpus runs anteroposteriorly along the inferior one-third of the medial face. The lower molars are damaged, eroded and moderately to heavily worn. M1 has been almost entirely obliterated by weathering and rolling, leaving only a well-rounded dentine core and a small fragment of enamel, 4 mm long, along the distal margin of the crown. M2 has a rounded mesiolingual face that has been eroded and rolled.
7 Hominins from Laetoli
A large crack passes transversely across the crown at an oblique angle, passing through the hypoconid and entoconid. The enamel rim from the buccal margin of the protoconid has been chipped away. The M3 is complete, but several cracks have caused slight expansion of the dimensions of the crown. Mesially there is a transverse fracture that passes across the mesial aspect of the protoconid and metaconid to skirt the distal margin of the mesial fovea. A second fracture originates from the midpoint of the transverse crack and passes distally in the midline into the center of the talonid basin, where it deviates buccally to pass through the distal margin of the hypoconid. A third fracture passes mesiolingually across the mesiobuccal face of the hypoconulid at an oblique angle to join the midline fracture. A finer crack passes between the hypoconulid and entoconid. The bone of the mandibular corpus is yellow-grey in coloration with a black tar-like staining, the enamel of the teeth is dark grey with orange staining, and the exposed dentine is black. This resembles the coloration of fossils from late Pleistocene sediments, but given the degree of weathering and rolling, the specimen clearly spent a long period of time on the surface prior to being recovered, and it is very likely that the fossil developed a coloration and patination associated with the rare fossil material derived from the surface lag.
Morphology The mandibular corpus is poorly preserved. It is robust, being thickest in the alveolar region and narrowing inferiorly, and increases in thickness posteriorly. The corpus maintains a relatively constant depth below the molars or deepens slightly posteriorly. The lateral face of the corpus is slightly convex inferosuperiorly. The root of the ramus originates opposite M3. The medial face is relatively flat. The M1 is poorly preserved; it comprises a rounded stump of dentine, with a small segment of the enamel rim preserved along the distal margin. The occlusal surface of the enamel remnant is worn flat, indicating that the molar was probably very worn with
157
much of the enamel cap missing. The dimensions of M1 cannot be measured or accurately estimated, but preservation of the enamel distally and the mesial root does indicate that the tooth was at least 10 mm in length. The M2 is mostly complete, but the mesiolingual corner and much of the lingual face have been damaged and smoothed as a result of erosion and rolling. The length of the crown can be measured with precision, but only a minimum and estimated buccolingual breadth can be obtained (see Table 7.8). The crown appears to have been mesiodistally longer than broad and rectangular in shape with rounded corners. The enamel cap is worn flat and large areas of dentine are exposed on the protoconid and hypoconid and a smaller area is exposed on the hypoconulid. Damage to the lingual side of the crown does not permit determination of the extent of dentine exposure on the metaconid and entoconid. The crown is too worn and damaged to provide any information on the detailed morphology of the occlusal surface. In terms of its size, proportions, cusp distribution, and enamel thickness it appears to be consistent in morphology with other M2s of A. afarensis (see below). The M3 is complete, but the crown is cracked and moderately worn. The crown is longer than broad and ovoid in shape, with well-rounded corners. The mesial margin of the crown has a broad and concave interproximal facet for contact with M2. The length of the tooth is currently 14.1 mm, but its original length can be estimated to have been 14.8 mm (Table 7.8). M3 was clearly larger in area than M2. Based on the estimated dimensions of the molars, the occlusal area (mesiodistal length ´ buccolingual breadth) of M3 is 9.8% larger than that of M2. The protoconid is larger than the hypoconid, which is situated immediately distally. The hypoconulid is a small triangular cusp situated just to the buccal side of the midline of the crown. The three buccal cusps are closely approximated, being separated by fine fissures that radiate out from the talonid basin. The metaconid is smaller in area than the protoconid. A short groove running obliquely along the mesiobuccal margin of the metaconid represents the remnant
Table 7.8 Dimensions (mm) of lower molars in L.H. 29 compared with other A. afarensis teeth from Laetoli, Hadar, and Maka Laetolib Hadarc Makad a Tooth Dimension L.H. 29 N Mean Range N Mean Range N Mean Range M2
MD 13.2 (13.8) 4 14.3 14.0–14.9 25 14.2 12.1–16.5 3 14.4 14.0–15.0 BL (12.9) 4 13.4 12.7–14.1 21 13.4 11.1–15.2 3 13.2 13.0–13.3 M3 MD 14.1 (14.8) 1 16.2 16.2 19 15.1 13.4–17.4 4 15.6 15.2–16.2 BL 13.4 1 13.9 13.9 16 13.4 11.3–15.3 4 13.4 13.0–13.8 MD mesiodistal length, BL buccolingual breadth a Estimated values are in parentheses. Data collected by the author b Specimens collected by Mary Leakey (1974–1979), excluding L.H. 29. Data collected by the author c Data from Kimbel et al. (2004) d Data from White et al. (2000). The data include estimated and maximum values, as well as teeth from both sides of the MAK-VP-1/12 mandible
158
of the mesial fovea. The entoconid is similar in size to the metaconid. Interposed between the metaconid and entoconid is a small wedge-shaped metastylid, separated from the main buccal cusps by deep fissures. A V-shaped distal basin separates the entoconid from the hypoconulid. A small area of dentine is exposed on the protoconid, and the other buccal cusps are worn flat but without exposed dentine. The metaconid and entoconid exhibit a moderate degree of wear, but they retain greater topographic relief than the buccal cusps. Bordering the buccal face of the protoconid and the mesiobuccal face of the hypoconid is a rounded, but welldeveloped buccal cingulum (or protostylid, expression state 6 of Hlusko 2004), as is typically found in the lower molars of A. afarensis from Laetoli. The cingulum is better developed than on L.H. 4, but is comparable to that seen on M1 and M2 of L.H. 14h. Comparisons The height and breadth dimensions of the mandibular corpus below mid-M2 can be compared with data on A. afarensis mandibles from Hadar (see Table 7.9). The breadth of the corpus in L.H. 29 is 22.8 mm, compared with a mean value of 21.8 mm in the sample from Hadar (range = 18.1-30.5 mm; n = 24) (Kimbel et al. 2004). The height of the corpus at M2 is 38.4, which exceeds the deepest mandible known from Hadar (A.L. 444-2). The mean corpus height at M2 in the Hadar sample is 31.7 mm (range = 25.3–37.6; n = 19) (Kimbel et al. 2004). L.H. 29 is similar in corpus breadth to L.H. 4, but is somewhat deeper (Table 7.7). The size of the corpus indicates that L.H. 29 was probably a male individual. It also indicates that there was a good deal of variation in corpus dimensions in the Laetoli sample, comparable to that from Hadar. If L.H. 29 does derive from the Upper Laetolil Beds, then it may have important implications for inferences about evolutionary change in A. afarensis. Lockwood et al. (2000) have demonstrated a shift toward larger corpus size in the youngest A. afarensis mandibles from Hadar, but the recovery of an individual with a relatively large mandibular corpus from Laetoli, earlier in time than the Hadar sample, may call into question whether this was a temporal trend in the species as a whole or a more localized phenomenon uniquely characteristic of the Hadar sample. The M1 and M2 are too worn and damaged to allow detailed comparisons of the occlusal morphology. In its shape, proportions, cusp size and distribution, and enamel thickness, the M2 in L.H. 29 matches well with the corresponding tooth in L.H. 4, but is slightly smaller. In terms of its estimated occlusal area (178.2 mm2) it represents the smallest M2 so far recovered from Laetoli (previous finds range in occlusal area from 180.3 to 197.4 mm2). However, the estimated occlusal area does fall well within the range of
T. Harrison Table 7.9 Comparison of dimensions (mm) of the mandibular corpus in L.H. 29 with other specimens of A. afarensis Locality Specimen Ht at M2 Br at M2 Laetoli
L.H. 29 38.4 22.8 L.H. 4 30.3 22.6 L.H. 13 – 23.6 Hadar A.L. 128-23 – 22.9 A.L. 145-35 – 24.8 A.L. 188-1 34.3 18.8 A.L. 198-1 30.8 18.1 A.L. 198-22 34.0 20.9 A.L. 207-13 25.3 18.4 A.L. 225-8 28.1 21.4 A.L. 266-1 27.6 24.2 A.L. 315-22 28.0 20.0 A.L. 330-5 28.3 19.5 A.L. 333w-1a, b 32.4 23.0 A.L. 333w-32 + 60 35.4 23.6 A.L. 417-1a 32.8 18.4 A.L. 432-1 – 20.3 A.L. 433-1a, b – 20.8 A.L. 436-1 26.0 19.6 A.L. 437-1 – 19.6 A.L. 437-2 37.0 24.2 A.L. 438-1 g 37.1 28.1 A.L. 444-2 37.6 30.5 A.L. 620-1 34.5 22.6 Maka MAK 1/12 29.6 20.6 MAK 1/2 32.6 21.4 a Dimensions: Inferosuperior height of corpus at mid-M2; mediolateral breadth of corpus at mid-M2 b Data: Laetoli (Harrison, unpublished); Maka (White et al. 2000); Hadar (Kimbel et al. 2004). Where both sides are measurable, the value is the average of the right and left sides
variation for the Hadar sample (n = 21, range = 137.6– 234.1 mm2, mean = 189.0; Kimbel et al. 2004). The M2 crown is relatively narrow with an estimated breadth-length index of 93.5, which falls within the range of variation of the previously collected specimens from Laetoli (n = 4, range = 89.4– 100.7, mean = 93.6) and the Hadar sample (n = 21, range = 84.6–107.7, mean = 95.6; Kimbel et al. 2004). The M3 in L.H. 29 is similar to L.H. 4, the only other M3 previously recovered from Laetoli (L.H. 15 is considered here to be an M2 rather than M3, contra White 1980a), but it differs in being smaller in size (the occlusal area in L.H. 29 is 198.3 mm2 versus 225.2 mm2 in L.H. 4) and relatively broader (the breadthlength index in L.H. 29 is 90.5 versus 85.8 in L.H. 4), with more convex buccal and lingual margins, a larger hypoconulid, a larger distal fovea, a more pronounced metastylid, and a better developed buccal cingulum. The occlusal morphology of M3 in L.H. 29 corresponds well with examples from Hadar. The occlusal area and breadth-length index fall close to the mean values for the Hadar sample (occlusal area: n = 15, range = 151.4– 266.2 mm2, mean = 200.0; breadth-length index: n = 15, range = 82.4–100.7, mean = 89.0; Kimbel et al. 2004).
7 Hominins from Laetoli
In L.H. 29 M3 is slightly larger (11.4% larger) in its occlusal area than M2. A similar relationship occurs in L.H. 4 (17.4% larger than M2) and the Hadar sample (the occlusal area of M3 exceeds that of M2 in associated teeth by an average of 9.4%), but there is a good deal of variation in the latter sample, with some specimens having M3 slightly smaller than M2 (up to 9.0% smaller) or much larger (up to 50.4% larger) (Kimbel et al. 2004).
LAET 79-5447 Right upper canine (Fig. 7.7).
Location and Stratigraphic Provenance According to Mary Leakey’s field catalogue this specimen was found on July 6, 1979 at Loc. 8, but details of its stratigraphic provenance and the identity of the collector of the specimen are not recorded. Although identified as a “hominid” in the catalogue, it presumably remained undescribed and unpublished because of the weathering and dark staining, similar to L.H.29, which gives it a superficial resemblance to the preservation typical of late Pleistocene fossils at Laetoli (see below).
159
crack runs perpendicular, passing from the apex onto the buccal and lingual faces where it terminates at mid-crown. A series of smaller hairline cracks disrupt the enamel around the base of the crown, especially on the buccal aspect. The buccal face of the crown is eroded and pitted by weathering. The lingual face is more heavily etched, and much of the enamel from the base of the lingual aspect of the crown has been lost. A small chip of enamel is missing from the mesial marginal ridge on the lingual face. The enamel in the unweathered state was grey-brown, as is common among fossil teeth from the Upper Laetolil Beds, but most of the enamel surface has developed a white patination, with orange staining and smaller areas of black patination. The root was originally grey-brown, but is now stained black and orange with light grey patches. This black and orange mottling is typical of Late Pleistocene fossils, and this probably accounts for why the specimen has previously been overlooked. However, as noted above for L.H. 29, specimens that erode out of the Upper Laetolil Beds and become reworked with the superficial sediments often secondarily develop a patination and stain resembling Late Pleistocene fossils. The original coloration of the fossils does indicate that it was derived from the Upper Laetolil Beds, and this is supported by the anatomy and size of the tooth, which is identical to that of upper canines of A. afarensis (see Table 7.10).
Morphology Preservation The specimen preserves a complete crown and most of the root, except for the tip. It is quite heavily weathered. The crown has a prominent crack that runs from the apex along the mesial ridge and margin and continues along the mesial face of the root to terminate at the broken root tip. A similar
The entire crown and most of the root are preserved. The crown is relatively tall, with an angular apex. The apicomesial margin runs for 6.5 mm from the apex to form a long and relatively sharp ridge. It is angled at 23° to the apicobasal axis of the crown. Light wear is evident along the mesiolingual aspect of the apex, where a small area of dentine is exposed, and it continues mesially as a flattened facet along the apicomesial margin of the crown. This wear facet is produced by occlusion with the apex of the lower canine. The apicomesial margin meets the mesial margin at an obtuse angle of 126°, and forms the mesial shoulder situated about two-thirds down from the base of the crown. From the mesial Table 7.10 Dimensions (mm) of upper canines of Australopithecus afarensis Laetolia Hadar and Dikikab Dimensions LAET 79-5447 N Mean Range N Mean Range
Fig. 7.7 LAET 79-5447, right upper canine of Australopithecus afarensis. (a) distal view; (b) mesial view; (c) buccal view; (d) lingual view
MD 10.3 3 10.4 9.5–11.7 12 9.7 8.8–10.4 BL 9.3 3 10.0 9.1–11.5 11 10.9 9.3–12.4 BHT 14.2 2 14.0 13.9–14.0 MD mesiodistal length, BL buccolingual breadth, BHT buccal height of crown a Data collected by author b Data from Kimbel et al. (2004) and Alemseged et al. (2006)
160
shoulder, the mesial margin ascends at a slightly oblique angle towards the base of the crown. The distal margin forms a relatively straight and sharp ridge that meets the apicomesial margin at an angle of 82°. Basally, the distal margin ridge terminates at a small distal tubercle, forming the distal shoulder, situated about one-third down from the base of the crown. The buccal face of the crown is apicobasally and mesiodistally strongly convex and generally featureless. The distal tubercle is bordered mesially by a shallow apicobasally oriented groove. A very shallow groove is also evident bordering the mesial margin. Despite the weathered enamel there is no evidence of any developmental disturbances, such as hypoplasias. The mesial face of the crown forms a flattened V-shaped surface. The enamel junction at base of the mesial aspect of the crown is U-shaped. Although the mesial face is weathered and eroded, traces of the original enamel surface confirm that the base of the crown had a large interproximal contact facet for I2 and that there was no maxillary diastema (absent in 50% of A. afarensis specimens according to Kimbel et al. 2004). In lingual view, the mesial shoulder is much more apically placed than the distal shoulder, producing a rhomboidal shaped crown with a strong mesiodistal asymmetry. The lingual face is mesiodistally slightly convex. An eroded remnant of a narrow and sharp lingual pillar is preserved in the middle one-third of the crown. It presumably ran from the base of the crown to the apex in its original state. A small subsidiary ridge originates from the base of the crown, close to the origin of the lingual pillar, and passes obliquely mesially across the lingual face of the crown to join the distal ridge midway along its length. Mesial to the lingual pillar, the lingual face is generally quite flat, although some fine crenulations in the weathered enamel surface may indicate that there was originally some secondary wrinkling. A shallow groove and a raised rim border the mesial margin of the lingual face. A narrow and shallow groove separates the lingual pillar from the distal margin. The base of the crown of the buccal face is damaged, so the morphological details are not discernable. The root is almost complete except for the missing apex. It is quite short. The preserved portion is 14.8 mm long, but it was probably 18.5 mm long when complete. The root is mesiodistally compressed, and its maximum cross-sectional dimensions are 10.3 mm by 6.9 mm. The mesial face of the root is flattened, with a shallow groove running along its length. The buccal face is convex and thicker than the lingual face, giving the root a triangular cross-section.
Comparisons LAET 79-5447 is very similar in overall size and crown height to the unerupted upper canine, L.H. 6b from Laetoli,
T. Harrison
which is inferred to have belonged to a male individual. L.H. 6b does differ, however, in a few features, including: (a) a steeper and more convex apicomesial ridge that grades smoothly into the mesial margin, without the sharp angulation seen in LAET 79-5447; (b) a slightly more prominent lingual pillar; (c) a less pronounced distal tubercle; and (d) the mesial margin of the lingual face lacking the distinct marginal rim. The largest of the upper canines from Laetoli, LH 3e, is quite a bit larger than LAET 79-5447. The mesiodistal length and buccolingual breadth of the crown in LAET 79-5447 are 13.6% and 23.7% smaller than in L.H. 3e, respectively, but the crown height was slightly higher. LAET 79-5447 also differs from L.H. 3e in the following features: (a) the apicomesial ridge and the mesial margin are angled as in LAET 79-5447, but the angulation occurs at mid-crown height, rather than onethird up from the apex as in LAET 79-5447; (b) the lingual pillar is better developed and positioned closer to the mesial margin; (c) the distal margin is shorter and less steep, giving the crown in L.H. 3e greater mesiodistal symmetry in lingual view; and (d) there is more strongly developed secondary wrinkling on the buccal face. LAET 79-5447 is larger and relatively higher crowned than the worn upper canine in L.H. 5, inferred to have belonged to a female individual, although it is generally similar in morphology. The root is longer and more slender in L.H. 5. The sample of upper canines of A. afarensis from Hadar (n = 11; Kimbel et al. 2004) exhibits metrical variation that can be interpreted as sexual dimorphism in overall size (i.e., cross-sectional area and crown height). Metrically, LAET 79-5447 falls in the lower end of the size range of canines inferred to belong to male individuals (being most similar in dimensions and morphology to A.L. 200-1) (Table 7.10). LAET 79-5447 is morphologically similar to A.L. 333x-3 in the shape and degree of symmetry of the crown, but somewhat smaller in its overall dimensions (see Fig. 7.8). The latter specimen is the largest of the upper canines from Hadar, and is reasonably interpreted as belonging to that of a male individual, but it is still smaller than L.H. 3e, which is the largest upper canine attributed to A. afarensis. A.L. 333x-3 also differs from LAET 79-5447 in the following features: (a) the apicomesial ridge is slightly steeper; (b) the lingual pillar is similar in development, but slightly more mesially placed; and (c) the distal tubercle is less prominent. These relatively minor differences are outweighed by the striking similarities. Kimbel et al. (2004: 207) suggest that the small sample of upper canines from Laetoli can be distinguished from those from Hadar in having more symmetrical crowns with a more cervically positioned mesial shoulder, being more similar in height to the distal shoulder. This is the case for L.H. 3e and possibly also for L.H. 5, but not for L.H. 6b or LAET 79-5447, which have the pattern typically found in the upper canines from Hadar. This can be considered a variable feature in the A. afarensis sample from Laetoli, and claims about the
7 Hominins from Laetoli
161 Table 7.11 Taxonomy and synonymy list of Paranthropus aethiopicus Superfamily: Hominoidea Gray, 1825 Family: Hominidae Gray, 1825 Subfamily: Homininae Gray, 1825 Tribe: Hominini Gray, 1825 Genus: Paranthropus Broom, 1938 Species: P. aethiopicus (Arambourg and Coppens 1968) Synonymy 1967 – “Paraustralopithecus aethiopicus” – Arambourg and Coppens (1967) [unavailable, conditionally proposed] 1968 – Paraustralopithecus aethiopicus Arambourg and Coppens, 1968 – Arambourg and Coppens (1968) 1978 – Australopithecus africanus Dart, 1925 – Howell (1978) [partim] 1986 – Australopithecus aethiopicus (Arambourg and Coppens 1968) – Walker et al. (1986) 1988 – Paranthropus aethiopicus (Arambourg and Coppens 1968) – Clarke (1988) 1989 – Australopithecus walkeri Ferguson, 1989 – Ferguson (1989)
Fig. 7.8 Comparison of LAET 79-5447 with upper canine A.L. 333x-3 from Hadar. (a) A.L. 333x-3 (cast), lingual view, left upper canine (reversed image). (b) LAET 79-5447, lingual view, right upper canine
purported difference between the two samples have been influenced by the extreme morphology represented by the very large canine, L.H. 3e, from Laetoli. Nevertheless, it is true that the configuration seen in L.H. 3e and to some extent L.H. 5 is not matched by any specimens in the large sample of upper canines from Hadar, thereby implying that a discernable difference may exist in the relative frequency of different upper canine morphs between the samples from Laetoli and Hadar.
candidate taxa in eastern Africa were penecontemporaneous (i.e., Paranthropus aethiopicus, Australopithecus garhi, and Homo sp.). In 2001 a maxillary fragment was recovered from the Upper Ndolanya Beds at Silal Artum that was identifiable as Paranthropus aethiopicus. This is the only specimen definitively attributed to this species from outside the Turkana Basin of northern Kenya and southern Ethiopia (see Table 7.11 for a summary of the taxonomy and nomenclature of this species). With this discovery, it becomes more likely that the proximal tibia from Loc. 22S belongs to the same species, but the possibility that it represents a second hominin taxon that is not yet identified in the Upper Ndolanya Beds cannot be entirely ruled out. As a consequence, it is left unattributed. In addition, a possible hominin, previously undescribed, was recovered by Mary Leakey’s expeditions in 1975 during excavations in the Upper Ndolanya Beds at Loc. 7E. The specimen consists of a small fragment of the right zygomatic process of the frontal bone of an infant, associated with two other indeterminate cranial fragments. The specimen was originally identified as a cercopithecid, but the size and morphology are consistent with it being a hominin. The specimen is provisionally identified here as cf. Hominini indet.
EP 1500/01 New Hominins from the Upper Ndolanya Beds Since the resumption of systematic fieldwork at Laetoli in 1998 fossil hominins have been recovered for the first time from the Upper Ndolanya Beds, which are dated to 2.66 Ma. In 1998 a proximal tibia (EP 1000/98) was discovered at Nenguruk Hill at Loc. 22S. At the time, its taxonomic affinities could not be determined because hominin craniodental specimens were not known from the Upper Ndolanya Beds, and several potential
Edentulous right maxilla with the roots for I2-M2 (Fig. 7.9). Location and Stratigraphic Provenance The author discovered the specimen on July 28, 2001 at the locality of Silal Artum. The specimen was a surface find, partially reburied, but, along with the associated fauna, was unquestionably derived from the Upper Ndolanya Beds.
162
T. Harrison
7 Hominins from Laetoli
Detailed information on the locality and hominin find spot are presented in Harrison and Kweka (2011).
Preservation The specimen consists of a right maxillary fragment comprising a large part of the maxillary corpus. The specimen is edentulous, but preserves the roots of I2-M2. The specimen was found partially buried in reworked sediment resting with the roof of the palate facing upwards. There is a sharp change in coloration, from creamy white to orange, across the lateral face of the maxilla that delineates the portion that was buried (orange) from that portion exposed to the elements. The specimen is lightly rolled and abraded. The anterior part of the maxilla containing the upper canine and incisors is missing. Judging from the sharp break in the bone and the fresher appearance of the exposed surfaces of the roots of the canine and I2 relative to the cheek teeth, it would appear that this portion was sheared off subsequent to the specimen being weathered out onto the surface, and being rolled and abraded. Given the force needed to cleanly detach the bone fragment, it is likely that the specimen was damaged by trampling. From the orientation of the specimen at the time of discovery and from the color difference of the exposed portion of the maxilla, we know that the subnasal region of the maxilla was projecting above the soil surface and would have been prone to trampling. Nevertheless, the occurrence of adhering matrix to the broken surface of the bone does indicate that some time had elapsed between when the damage was done and when the specimen was collected. The detached piece was not recovered during screening operations following initial discovery of the specimen. The alveolus for I1 is matrix filled, indicating that this tooth had been lost before fossilization, but because the alveolar region of the maxilla is missing it is not possible to determine whether the tooth was lost antemortem or postmortem, but the latter seems the most likely scenario. Posteriorly, the palate is preserved as far as the anterior alveolar wall of M3, but there is no trace of a tooth root. Matrix adhering to the posterior wall of the maxilla at the time of discovery confirms that M3 was already lost at the time of fossilization. Nevertheless, the contour of the anterior alveolar wall confirms that M3 was erupted at the time of death. The surface bone of the palate exhibits numerous fine cracks, and these appear to be largely or exclusively the result of postfossilization weathering. The palate is preserved as far as the midline from opposite the mesial root of P3 to opposite mid-M2. Posteriorly the midline region of the palate has been rounded by rolling and abrasion, and most of the horizontal plate of the palatine
163
bone is missing medially and laterally opposite M3. The lateral margin of the alveolar process of the maxilla has been slightly abraded and smoothed by rolling, especially at the level of the posterior root of M1 and M2. This gives the alveolar region and exposed roots a strongly convex contour mediolaterally. The pulp cavity chambers of the cheek teeth have a fine film of calcite covering them, which indicates that some time had elapsed between the teeth being detached from the maxilla and the time of its discovery. Laterally the root of the zygomatic process is largely missing and the edges of the broken bone are well-rounded, especially inferiorly and posteriorly. On the lateral aspect of the lower face the surface bone is weathered and pitted by weathering. Small flakes of bone have been lost from the alveolar region to expose the lateral aspect of the roots of most of the cheek teeth. This damage most likely occurred prior to fossilization when the bone was still fresh. Facially, the frontal process of the maxilla is lacking, but the inferolateral margin of the nasal aperture is preserved. No portion of the orbit or orbital wall is preserved. At the time of discovery the maxillary sinus and the inferior meatus of the nasal cavity were filled with sediment. The matrix has since been removed to expose the morphological details of the sinus and nasal passage. Based on the damage and weathering it is possible to develop a scenario of the series of events that took place in the preservation of the specimen. The maxilla was separated from the rest of the cranium prior to burial and exposed on the Pliocene land surface long enough for weathering to produce some cracking and pitting of the bone surface. During this period, the maxillary fragment was detached from its counterpart on the left side, and the fragile bone covering the lateral roots of the cheek teeth was damaged and flaked away. I1 and M3 dropped out of their respective alveoli and were lost. Subsequently, the specimen was buried by ashes of the Upper Ndolanya Beds and fossilized. On weathering out of the sediment, the specimen was exposed to further weathering and experienced some degree of tumbling and transportation that caused rolling and abrasion of the maxilla and loss of the remaining teeth. The subnasal region of the maxilla was detached at a later point in time, probably as a result of trampling, as it lay partially buried upside down in the unconsolidated soil in the location where it was found.
Morphology The palatal region of the maxilla is preserved to the midline suture from opposite P3 to mid-M2. The incisive canal was filled with matrix at the time of discovery, but has since been
Fig. 7.9 EP 1500/01. Right maxillary fragment of Paranthropus aethiopicus from Silal Artum. (a) anterior view; (b) lateral view; (c) superior view; (d) medial view; (e) inferior view
164
cleaned. Just posterior to the anterior broken edge of the maxilla, opposite the posterior margin of the canine root, is a shallow groove that represents the vestibule of the incisive foramen. This groove passes posteriorly for a distance of 11.2 mm, where it disappears beneath the roof of the palate to enter the incisive canal transversely opposite the mesial margin of P4. The groove deepens as it passes posteriorly, and is angled slightly (~6°) towards the midline. The canal itself forms a bilaterally paired structure, partially separated by a raised keel that passes along the midline of the floor of the nasal canal and the midline of the floor of the inferior surface of the clivus. The dorsal surface or roof of the canal has a low and fine keel that runs posteriorly midway along its length. In section, the palatal aperture of the incisive canal is elliptical in shape with a dorsoventral height of 3.1 mm and a breadth to the midline of 4.1 mm. The vestibule and canal combined have a minimum anteroposterior length of 21.4 mm, measured along the midline of the clivus. In superior view, the canal is funnel shaped, becoming wider as it passes posteriorly, and it eventually opens into a broad and shallow groove running along the floor of the nasal passage. The incisive canal is directed posteriorly at a 30° angle relative to the roof of the palate. In medial view the nasoalveolar clivus overlaps the hard palate (see McCollum 1997). The palate is relatively wide, with a slight increase in breadth posteriorly. The estimated breadth of the palate at mid-P3, measured externally to include the alveolar process, is 66.2 mm, whereas at mid-M2 it is 81.4 mm. The palatal breadth between the roots can be estimated to have been 32.6 mm between the mid-P3s and 44.6 mm the between mid-M2s. The depth of the palate is difficult to measure with precision because it is not a simple task to orient the maxilla in the correct mediolateral plane. However, by aligning the alveolar process and the midline axis with the parasagittal plane, it is possible to approximate the correct anatomical position. The palate is flat and relatively shallow, with a depth of only 15 mm in the midline at P3/P4 and appears to retain a constant depth posteriorly. The palate is bordered laterally by the sloping internal wall of the alveolar process, which increases in steepness posteriorly. Beginning anteriorly opposite mid-M1 is a shallow groove for the greater palatine vessels. Posteriorly opposite M2 this develops into a pair of grooves, 3.8 mm wide, separated by a low keel. The keel terminates at mid-M2 to become a shallow single groove after which it begins to ascend more steeply as it approaches the entrance of the greater palatine canal. No portion of the greater palatine foramen is preserved, but it was presumably located opposite M3. A section of the palatomaxillary suture is visible on the palate originating ~11 mm lateral to the midline opposite anterior M2. It passes laterally and slightly posteriorly to terminate at the lateral margin of the groove for the greater palatine vessels. The point at which the palatomaxillary suture intercepts the intermaxillary suture in the midline is not preserved.
T. Harrison
The sheared-off subnasal clivus exposes the alveolus for I1 and roots of the right I2 and upper canine. The alveolus for I1 is filled with matrix, implying that the tooth was lost perimortem prior to fossilization. Since the clivus and the roots of the canine and incisors were broken obliquely, the crosssectional area of the roots are greater than they would be if they were broken transversely. Nevertheless, it is possible to compare the size of the roots of the two incisors and to estimate the original diameter of the canine root at the point of breakage. The size of the incisor roots/alveoli indicates that I1 and I2 were subequal in size, and relatively small compared to the size of the roots of the cheek teeth. The apico-basal orientation of the roots and the configuration of the subnasal region indicate that the incisors were procumbently implanted in the premaxilla. The intercanine distance is estimated to be ~29 mm, while the external breadth of the palate at the level of the canine roots can be estimated to have been ~50 mm. The I1 alveolus shows that the root was ovoid in cross-section with a buccolingual diameter much greater than the mesiodistal diameter. A portion of the I2 root is retained, but it appears to be broken close to the apex of the root. The I2 and canine root are separated by a distance of only 1.6 mm, indicating that there was no diastema. The canine root is pyriform in section, narrowing mesiobuccally, and with a concavity on the mesiolingual side. The dimensions of the broken canine root are 11.8 mm (mesiodistal length) and 14.1 (buccolingual breadth), but these are maximal dimensions because of the obliquity of the break. By taking the measurements orthogonal to the longaxis of the root it is possible to estimate that the root was ~10 mm long and ~13 mm wide. This indicates a large canine root, and presumably a large crown, in absolute terms, but one that is relatively small in terms of the overall size of the palate and the roots of the cheek teeth. The lateral margin of the canine root is set medial to the lateral margin of P3, so the latter tooth forms the anterolateral corner of the rectangular palate. The root of the canine is procumbently implanted at an angle of about 45° to the anteroposterior plane of the palate. The root appears to be relatively straight, with only a slight degree of apicobasal curvature. None of the crowns of the cheek teeth is preserved, but the roots of P3-M2 are exposed in the alveolar process. The roots indicate that the cheek teeth were aligned in relatively straight line that diverged slightly posteriorly. The length of the P3M2 chord, based on the roots alone, is 49.3 mm, although with the reconstructed length of the crowns it can be estimated to have been ~51 mm. This implies that the cheek teeth were absolutely very large. The premolar roots are extremely broad suggesting that the anterior cheek teeth were massive. The section of the roots for P4 is slightly larger than that for P3. The bases of the roots are ovoid in outline with single pulp chambers, but these divide into buccal and lingual canals that enter separate
7 Hominins from Laetoli
roots. The buccal root of P3 is prominent, causing the lateral wall of the maxilla to bow outwards slightly, and it curves slightly medially to converge with the canine root. The buccal root in P4 is partially visible on the lateral side of the maxilla where a small flake of bone has been lost from the lower face. The exposed root is angled slightly anteriorly relative to the midline transverse axis of the pulp cavity. From the external morphology it is likely that the buccal roots of P3 and P4 are paired, but this will need to be confirmed with x-ray or CT scanning methods. The pulp cavities of M1 and M2 are relatively large. Each exposes three canals, which lead to separate roots. The lateral roots are partially exposed. The mesiobuccal root penetrates the alveolar process of the maxilla almost vertically, while the distobuccal root is angled posteriorly with a slight distal curvature. The lingual root is single, and judging from the base of the root, the medial face is strongly grooved to produce a bilobate cross-section. Posterior to M2 is a short section of the lamina of bone that separates it from the alveolus of M3. The anterior face of the M3 alveolus has a sharp inferior border, indicating that the last molar was erupted and in place at the time of death. Apart from concluding that the specimen belonged to an adult individual, no other particulars of its age can be given. Judging from the size of the canine root and from the overall robusticity of the maxillary fragment it is possible to infer that the specimen belonged to a male individual. Anteroposteriorly, the lateral alveolar wall forms a straight line, rather than bowing laterally around the molar row, and diverges slightly posteriorly (5°) relative to the midline axis of the maxilla. This indicates that the palate was rectangular, rather than forming a laterally bowing palate as seen in A. afarensis. The lateral surface of the maxilla in the alveolar region of the molars is relatively vertical. There is a slight protuberance of the alveolar region anteriorly to accommodate the buccal root of the P3. Posterior to this is a small, shallow fossa located above and posterior to the region of P4. Anterior to the premolars, the alveolar region of the maxilla curves sharply around the root of the canine. There is no distinct anterior pillar or canine jugum associated with the canine root, just a low rounded eminence. There is a very shallow canine fossa (i.e., the maxillary fossula) between the area of the maxilla filled by the canine root and the zygomaticoalveolar crest (Rak 1983, 1985). The curvature of the maxilla around the canine root implies a sharp demarcation between the lateral and anterior regions of the face, and a relatively abbreviated muzzle anteriorly. The facial region anterior to the zygomatic process and bordering the nasal aperture is very slightly concave mediolaterally, to produce a dished mid-facial region (see Rak 1983, 1985). The lower face slopes steeply down from the infraorbital region, at an angle of 46° to the palatal plane (i.e., the nasocanine angle), to the horizontal level of the
165
middle of the nasal aperture, after which it descends to the incisal region slightly less steeply, at an angle of 38° (i.e., the nasoalveolar angle). Although much of the region inferior to the nasal aperture is not preserved, the configuration of the face lateral to the nasal aperture suggests that subnasal prognathism was pronounced. The minimum dorsoventral thickness of the broken surface of the subnasal alveolar region at I1 is 12.2 mm in EP 1500/01. A similar thickness occurs at a distance of 10.6 mm posterior to prosthion in KNM-WT 17000, so we can estimate that at least 10 mm is missing from the front of the palate in EP 1500/01. If this is the case, then the anteroposterior length of the subnasal clivus (prosthion-nasospinale) can be estimated to have been ~37 mm. The root of the zygomatic process is very thick anteroposteriorly, with a broad and smoothly rounded inferior margin. The anterior margin of the zygomatic process is positioned above mid-P4 and the mid-point of the process is positioned above mid-M1. It has a minimum height above the alveolar margin of ~15 mm. The anterior face of the zygomatic process is relatively flat to slightly concave mediolaterally, and flares laterally at an angle of 80° to the long-axis of the lateral margin of the alveolar process in inferior view. The low rounded zygomaticoalveolar crest descends the lower face obliquely from the anteroinferior margin of the zygomatic process to terminate just posterior to the P3 root. This forms the anterior margin of the shallow depression in the maxilla above P4, and demarcates the lateral face of the maxilla from the anterior face. In anterior view, the zygomaticoalveolar crest is weakly arched. Superior to the infraorbital foramen the maxilla is slightly mediolaterally concave, but this transitions to a slightly convex surface as it approaches the nasal aperture. This convex surface between the infraorbital foramen and the nasal aperture extends inferiorly to become continuous with the eminence for the canine root, but it does not form a distinct anterior pillar. The center of the infraorbital foramen is located 31.5 mm from the midline in the mediolateral plane and 17.6 mm from the margin of the nasal aperture, opposite the aperture’s greatest width. The inferior margin of the foramen is horizontally in line with the anterior nasal spine. It is positioned dorsoventrally above P4 at a height of 35.9 mm above the alveolar margin. The distance between the center of the foramen and the inferior margin of the zygomatic process is 17.2 mm. The foramen is situated quite low on the face, midway between the nasal aperture and zygomatic process. The infraorbital foramen is circular in outline, with a mediolateral breadth of 4.3 mm. The foramen opens inferiorly and slightly laterally into a distinct teardrop shaped depression. Unfortunately, no portion of the inferior orbital margin is preserved, so the lower facial height cannot be measured or estimated, but based on the preserved anatomy the lower face was evidently relatively deep. The minimum height of the
166
lower face above P3 is 48.8 mm to the point where the maxilla is broken superiorly. Although only a portion of the nasal aperture is preserved, it was clearly pyriform in outline, being broadest inferiorly and narrowing superiorly. The lateral margin of the nasal aperture is smoothly rounded inferiorly, but a low crest is developed about 12.3 mm up from the floor of the nasal aperture about 3.9 mm below the point at which the maxillary fragment is broken superiorly. The greatest breadth of the nasal aperture, at the point where the lateral crest originates, is estimated to be ~28 mm. Laterally, the margin of the nasal aperture is bordered by the canine root, with the root of the lateral incisor situated more medially. The inferior margin of the nasal aperture curves gently medially and inferiorly to reach its most inferior point midway between the greatest lateral extent and the midline, and then it ascends ~3 mm to reach the elevated anterior nasal spine, which is located in the midline and somewhat recessed. A low, rounded crest passes laterally and slightly anteriorly from the anterior nasal spine to demarcate the junction between the clivus and the floor of the nasal cavity. The anterior nasal spine is eroded, but was evidently quite prominent. The superior margin of the subnasal clivus is damaged. The bone covering the region between the apices of the upper central incisors is largely missing. During removal of the adhering sediment in this region, the cleaning process exposed a small elliptical and smooth-walled pocket of pneumatized bone (15.4 mm ´ 8.2 mm) that occupied the dorsal surface of the clivus on either side of the anterior nasal spine. Nevertheless, despite this damage, it is possible to infer that the subnasal clivus would have had a slightly convex dorsal margin anteroposteriorly that led smoothly into the floor of the nasal aperture without a distinct nasal sill (Robinson 1953; Ward and Kimbel 1983; McCollum et al. 1993; McCollum 1997). In anterior view, the inferior margin of the nasal aperture is mediolaterally concave, forming a nasoalveolar gutter (Rak 1983, 1985; Strait and Grine 2004). The floor of the nasal cavity has an elevated and mediolaterally bladelike spine in the mid line. It is clear that the vomer inserted along this crest anteriorly as far as the anterior nasal spine (Robinson 1953; McCollum 1997, 1999). The lateral wall of the cavity, close to the superior break in the maxilla, bears a longitudinal crest that represents the root of the inferior concha. It slopes inferiorly as it passes posteriorly. Anteriorly it is positioned about 15.5 mm above the floor of the nasal canal, just posterior to the internal opening of the incisive canal, whereas posteriorly it is positioned 10.2 mm above the floor of the nasal cavity at the level of M2 where the posterior break occurs. The lateral wall of the inferior meatus below the line for the attachment of the inferior concha is generally concave, although there is a distinct swelling superior and lateral to the internal opening of the incisive canal. The lateral wall of the nasal cavity separating the inferior
T. Harrison
meatus from the maxillary sinus is remarkably thick, with a maximum thickness along the broken margin of 6.4 mm, but it does get as thin as 1.6 mm. The palatal process of the maxilla is also very thick (McCollum 1997, 1999), with a maximum dorsoventral thickness of 8.4 mm opposite M1. The maxillary sinus is very extensive, extending laterally and anteriorly into the zygomatic process, posteriorly into the alveolar region beyond M2, and medially to border the nasal aperture. It does not penetrate the maxilla anterior to the zygomatic process in the region of the premolar roots, but it does extend anteromedially as far as the canine root. There is no evidence that it extended into the palatine process of the maxilla medially to form a recessus palatinus (Tobias 1967; McCollum 1997). The exposed floor of the sinus has a complex arrangement of buttresses and deep loculations between and around the roots of the molars. The buttresses primarily intersect each other at right angles, and the entire system is arranged so that the buttresses are oriented at ~45° to the mediolateral line of the zygomatic process. Presumably, the network of buttresses criss-crossing the floor of the sinus helps to maintain the structural integrity of the maxilla. A major buttress, more than 5 mm wide, originates at the medial wall of the sinus opposite the level of M2, and passes obliquely laterally and anteriorly to meet the lateral wall at a sharp vertically-oriented buttress, which coincide on the external surface of the maxilla with the posterior margin of the zygomatic process. Intersecting this transverse buttress is a longitudinally-oriented buttress that passes anteriorly and medially at an angle of about 90° to the former. It passes towards the base of the pillar of bone that represents the lateral margin of the nasal aperture, where it divides into short medial and lateral arms. This main buttress gives rise to smaller subsidiary buttresses midway along its length that pass medially and laterally. Posterior to the point where the main transverse and longitudinal buttresses intersect, there is a pair of sharp and diverging buttresses that pass to the posterior wall of the sinus. Between the buttresses, the floor of the sinus is excavated to form relatively deep loculi. Centrally, the floor of the sinus bears a very deep loculus bordered laterally and posteriorly by the main longitudinal and transverse buttresses respectively, and medially by the wall of the sinus. The anterior wall of the sinus bears the remnant of a bony canal that penetrates ~8 mm into the chamber of the sinus. This is the canal leading from the infraorbital foramen into the infraorbital groove that carries the infraorbital nerve and artery. The canal passes superiorly and slightly medially, which corresponds well with the orientation of the infraorbital foramen aperture on the facial aspect of the maxilla. Medial to the canal is a well-developed loculus that penetrates into the pillar that forms the lateral margin of the nasal aperture, just posterior to the canine root.
7 Hominins from Laetoli
Comparisons Comparisons with the crania of other Pliocene hominins indicate that EP 1500/01 exhibits a suite of derived features that it shares uniquely with Paranthropus spp. from southern and eastern Africa. These include: relatively small anterior teeth; enlarged premolars and molars; a dished midface; robust and anteriorly placed zygomatic process of the maxilla; zygomaticoalveolar crest weakly arched in anterior view; single infraorbital foramen positioned low on the face and opening inferiorly into a shallow groove; guttered nasoalveolar clivus that grades smoothly into the nasal cavity floor; marked overlap between the nasoalveolar clivus and the anterior margin of the hard palate; anterior vomer insertion that coincides with the anterior nasal spine; recessed anterior nasal spine; and very thick hard palate anteriorly (Rak 1983, 1985; Leakey and Walker 1988; Kimbel et al. 1988; McCollum et al. 1993; Strait et al. 1997; Suwa et al. 1997; McCollum 1997, 1999; Asfaw et al. 1999; Keyser 2000; Strait and Grine 2001, 2004; Steininger et al. 2008). Further comparisons show that EP 1500/01 is very similar in morphology to KNM-WT 17000 (Walker et al. 1986; Leakey and Walker 1988), and there can be no doubt that the maxilla fragment from the Upper Ndolanya Beds should be attributed to the same species, Paranthropus aethiopicus. EP 1500/01 shares the following distinctive (i.e., primitive) features with KNM-WT 17000 that are not found in P. robustus or P. boisei: larger anterior teeth, relatively flat and shallow palate, and pronounced subnasal prognathism. EP 1500/01 and KNM-WT 17000 are very similar. The nasal aperture and subnasal region appear to have been very similar in shape and configuration. In both specimens the nasal aperture is pear-shaped in outline, relatively broad low down (although the greatest breadth is somewhat higher in EP 1500/01), with a slightly concave inferior margin mediolaterally (i.e., a guttered nasoalveolar region) and a prominent anterior nasal spine. The maximum breadth of the nasal aperture is estimated to be ~28 mm in EP 1500/01 compared with 27.3 mm in KNM-WT 17000 (Leakey and Walker 1988). The inferolateral margin of the nasal aperture is rounded in both specimens, in contrast to the sharp lateral margin in A. afarensis. The crest that forms the lateral margin of the nasal aperture superiorly originates lower in KNM-WT 17000. The estimated depth of the subnasal clivus in EP 1500/01 is comparable to that in KNM-WT 17000, and in both specimens it is sagittally convex and grades smoothly into the nasal passage without a distinct nasal sill. In A. afarensis the clivus has a stepped nasal cavity floor with a strongly differentiated nasal sill (Kimbel et al. 2004). The midface in EP 1500/01 is similar to that of KNM-WT 17000 in lacking an anterior pillar, having a robust zygomatic process that is placed anteriorly relatively to the cheek teeth and being slightly dished. Mediolateral concavity of
167
the midface is relatively slight in EP 1500/01 and more pronounced in KNM-WT 17000. The lateral margin of the nasal aperture is bordered by a low rounded eminence in EP 1500/01, whereas in KNM-WT 17000 this region is relatively flat. Both lack canine juga. The I2 root in EP 1500/01 and KNM-WT 17000 is positioned medial to the lateral margin of the nasal aperture, as in other species of Paranthropus and A. africanus, but distinct from A. afarensis In lateral view, the contour of the lower face is very similar in KNM-WT 17000 and EP 1500/00. Both specimens appear to have strong subnasal prognathism. As described above, the lower facial and subnasal angles, relative to the anteroposterior plane of the palate, are 46° and 38° respectively in EP 1500/01, compared with 42° and 30° in KNM-WT 17000. The lower face in P. aethiopicus exhibits a similar overall angulation to that of P. boisei, except that the subnasal region is much more prognathic. The angulation of the lower face of A. afarensis is less steep (i.e., more prognathic), but the degree of facial protrusion is much more pronounced in the subnasal region in P. aethiopicus giving the latter taxon a greater degree of prognathism overall (Kimbel et al. 2004). The zygomatic process of the maxilla originates at approximately the same elevation in both specimens, with its lowest point vertically above mesial M1. The infraorbital foramina in KNM-WT 17000 are incompletely preserved, but they appear to have been located in a similar position to that in EP 1500/01 (the foramen is positioned slightly higher and more lateral relative to the nasal aperture in KNM-WT 17000). The foramen in KNM-WT 17000 is similar to EP 1500/01 in opening inferiorly into a shallow groove (Leakey and Walker 1988). The maxillary sinus in KNM-WT 17000 is infilled with matrix, so comparison with the sinus morphology in EP 1500/01 is not possible (Walker et al. 1986; Leakey and Walker 1988). Nevertheless, it is apparent that the sinus extended laterally into the zygomatic process and anteromedially into the lateral rim of the nasal aperture posterior to the canine alveolus (Leakey and Walker 1988). Leakey and Walker (1988) indicate that the sinus in KNM-WT 17000 extends medially into the palatine process of the maxilla, in contrast to EP 1500/01, which lacks a recessus palatinus. A portion of the floor of the maxillary sinus is exposed above M3 in KNM-WT 17000, and this confirms that, as in EP 1500/01, the maxillary sinus had a relatively thick bony floor above the molar roots. This contrasts with the maxillary sinus of A. afarensis seen in A.L. 200-1, which has a much thinner sinus floor. In addition, the medial wall of the maxilla separating the maxillary sinus from the nasal passage is relatively thick in P. aethiopicus, in contrast to the thin-walled sinus seen in A. afarensis. The palate in both specimens is relatively broad and very shallow (Leakey and Walker 1988; Suwa 1989). The breadth of the palate between the roots of P4 and M2 in KNM-WT
168
17000 is 41.3 and 35.7 respectively. These dimensions compare with estimated values of 38.8 and 39.2 in EP 1500/01. The palate in both specimens is similar in overall dimensions to that of OH 5 (P. boisei), but the palate in this latter specimen is much deeper. The palate becomes slightly wider and the tooth rows diverge posteriorly in EP 1500/01, whereas in KNM-WT 17000 the palate narrows slightly posteriorly (although Leakey and Walker (1988) consider that this is likely an artifact of preservation). The incisive canal has a similar configuration in EP 1500/01 and KNM-WT 17000. The canal exits onto the roof of the palate opposite mid-P4 and opens anteriorly into a fanshaped fossa that terminates just posterior to the alveolus of I1. The structure of the nasal septum and vomeral insertion is nearly identical in both specimens. The intercanine distance in KNM-WT 17000 is 36.0 mm, whereas the minimum estimated breadth is only 29.0 mm in EP 1500/01, implying that the upper incisors were probably larger in the former. By comparison, the intercanine distance in OH 5 is 31.7, implying that, just as in P. boisei, the incisors of P. aethiopicus were relatively small in relation to the size of the cheek teeth (Suwa 1989). A similar set of relationships emerges if the mediolateral breadth across the incisor roots is taken into account. In KNM-WT 17000 the breadth is 35.2 mm, compared with only 32.6 mm in OH 5, indicating that P. aethiopicus has larger upper incisors than P. boisei (see Leakey and Walker 1988; Suwa 1989). However, the same dimension in EP 1500/01 is estimated to be only 30.2 mm, indicating that there was variability in incisor size in P. aethiopicus, and that incisor breadth overlapped with that in P. boisei. The alveolus for the upper canine in KNM-WT 17000 measures 7.2 mm (mesiodistally) by 12.7 mm (buccolingually), which is smaller than the estimated dimensions of the canine root in EP 1500/01 (~10.0 mm by ~13.0 mm). Canine root size in KNM-WT 17000 is comparable to that in OH 5, which has a mesiodistal length of only 6.8 mm. The canine root in EP 1500/01 is much larger than in other Paranthropus specimens, and, in fact, exceeds the dimensions of the root of the largest canine of A. afarensis. A.L. 333x-3, a presumed male individual from Hadar, has basal root dimensions of 8.5 mm by 11.3 mm. Nevertheless, compared to the estimated size of the cheek teeth and the maxilla, the canine in EP 1500/01 would have been relatively smaller than in A. afarensis. It would seem that the anterior teeth were somewhat larger on average than in P. boisei, but as in the latter species they were small by comparison with the size of the premolars and molars. In addition, the canine roots in P. aethiopicus are more procumbently implanted and less apicobasally curved than in A. afarensis and P. boisei, which have shallower and less prognathic subnasal regions. Measurements of the P3-P4 chord and P3-M2 chord of the roots provide an estimation of the relative size of the cheek
T. Harrison
teeth. In KNM-WT 17000 these dimensions are 20.3 mm and 50.0 mm respectively, compared with 19.0 mm and 49.3 mm in EP 1500/01, indicating that that the teeth might have been slightly smaller in the latter. The same dimensions in OH 5 (P. boisei) are 19.6 mm and 50.3 mm respectively. Evidently, the cheek teeth in P. aethiopicus were as large (at least in their mesiodistal dimensions) as in P. boisei. Based on the roots, the cheek teeth in EP 1500/01 and KNM-WT 17000 formed a relatively straight line, unlike in A. afarensis where the cheek tooth rows bow laterally, being widest at the level of M2.
EP 1000/98 Left proximal tibia (Fig. 7.10).
Location and Stratigraphic Provenance The specimen was discovered by Chris Robinson on August 19, 1998 at Loc. 22 South. The specimen was found exposed on the surface at the base of Nenguruk Hill, a small hillock on the western side of the Olaitole River valley. The location of the find, as well as the lithology of the adhering matrix and the coloration of the specimen, confirms that it was originally derived from the Upper Ndolanya Beds. Additional details about the locality and provenance are presented by Harrison and Kweka (2011). Preservation The specimen consists of a left proximal tibia. Only a short section of the shaft is preserved, with a total proximodistal length of ~58 mm. The proximal articular surface is generally well preserved, but there has been some abrasion and weathering along the anteromedial and posterolateral margins. There are small depressions on the shaft anterior to the medial facet that look superficially like they may have been caused by bite marks from a small carnivore, but closer microscopic examination shows that they are shallow eroded pits caused by weathering. The posterior intercondylar depression, for the attachment of the medial meniscus and the posterior cruciate ligament, has been accentuated by erosion and by loss of the bone surface. The medial and lateral articular surfaces exhibit fine cracks around their perimeters, apparently caused by weathering when the bone was still green. Similarly, the heavy longitudinal cracks along the shaft, and the spiral and angular fractures at the broken end of the shaft, imply that the proximal end of the tibia was detached from the rest of the bone while it was still green and prior to it being buried and fossilized.
7 Hominins from Laetoli
169
Fig. 7.10 EP 1000/98. Left proximal tibia of cf. Paranthropus aethiopicus from Loc. 22S. (a) anterior view; (b) posterior view; (c) lateral view; (d) superior view; (e) medial view
Morphology The proximal end of the tibia is broad. The mediolateral breadth is 51.6 and the anteroposterior length is 33.8, with a breadth-length index of 65.5. The condyles are anteroposteriorly short in relation to their breadth. The medial condyle forms a D-shaped facet, with a central depression and a low rounded rim. The lateral condyle is elliptical in shape, with a slightly anteroposteriorly convex lateral surface and a more concave medial surface. The posteromedial margin of the lateral condyle is slightly elevated. Both condyles are mediolaterally concave, but the lateral condyle is more markedly so. The medial condyle is anteroposteriorly longer than the lateral condyle, but mediolaterally narrower. The medial
condyle measures 33.0 mm ´ 20.6 mm; the lateral condyle measures 26.6 mm ´ 22.1 mm. The surface area of the medial condyle is subequal in size to that of the lateral condyle (375 mm2 compared with 373 mm2). The long-axes of the two condyles converge posteriorly. Between the two condyles is a well developed and elevated intercondylar eminence, with prominent medial and lateral tubercles. The maximum height of the medial tubercle is 5.8 mm. The lateral tubercle is slightly lower. The eminence is located midway across the proximal end of the tibia in the parasagittal plane and slightly posterior of the midcoronal plane. Laterally, the articular surface of the medial condyle extends almost to the apex of the medial tubercle. In superior view, the medial tubercle forms an oblique crest that
170
continues anteromedially to become the raised lateral margin of the medial condyle. The lateral tubercle forms a robust crescent-shaped ridge that extends anteriorly and posteriorly to form the raised medial margin of the lateral condyle. The two tubercles are united by an elevated transverse crest, directed slightly posterolaterally. Anterior to the intercondylar eminence is a roughened triangular area, the anterior intercondylar area, bordered by the raised rims of the medial and lateral condyles. Within this area, there is a large triangular region that borders the medial condyle, defined anteriorly by a sharp crest that runs along the anterior margin and laterally by low swellings. This is the area for the anterior cornu of the medial meniscus. Lateral to this is a roughened area, perforated by numerous vascular foramina, that represents the site of the corpus adiposum infrapatellare and the attachment for the lateral meniscus. Posterior to these two areas, on the anterior slope of the intercondylar eminence, is a smooth triangular region for the attachment of the anterior cruciate ligament. The posterior intercondylar area has been damaged by erosion, so details of its morphology cannot be discerned. It is not possible, for example, to determine the shape of the posteromedial contour of the lateral condyle or whether or not the lateral meniscus had a posterior attachment. However, the area posterior to the lateral tubercle was evidently quite restricted implying that EP 1000/98 lacked a posterior cornu of the lateral meniscus. Posterior retroversion of the head is minimal. The articular platform is only very slightly posteriorly tilted in medial view in relation to the long-axis of the shaft. The medial condyle is tilted 4°, whereas the plane of the lateral condyle is tilted only 1°. The lack of retroflexion of the head is structurally associated with a weak overhang of the condyles posteriorly, a steep anterior margin, and a less prominent tibial tuberosity. Only a short section of the shaft is preserved. It is mediolaterally slightly compressed and triangular in section. The anteroposterior diameter of the shaft at the inferior margin of the tibial tuberosity is at least 21.2 mm, while its mediolateral breadth at this level is 20.2 mm. In anterior view, the shaft below the medial condyle is more robustly built than that below the lateral condyle, which is more hollowed out (Aiello and Dean 1990). This hollowing of the lateral side of the shaft may be more closely tied to the structural and allometric relationships between the size of the tibial plateau and the mediolateral diameter of the shaft, than to the development and disposition of the tibialis posterior muscle (contra Aiello and Dean 1990; Berger and Tobias 1996). The mediolateral breadth of the shaft at the level of the inferior margin of the tibial tuberosity is only 39.9% that of the total breadth of the tibial plateau. The remaining portion of the shaft appears to have a long-axis that is approximately in line with the sagittal midline of the articular surface (measurement of the
T. Harrison
midline of the shaft to the margin of the lateral condyle in the mediolateral plane is 24.3 mm or ~47% of the total breadth of the proximal end of the tibia), and it exhibits no discernable degree of torsion. The anterior face of the shaft bears a low and domed tuberosity for the attachment of the patella ligament. Its superior margin is situated 22.2 mm inferior to the margin of the anterior intercondylar area. The tuberosity projects anteriorly very little, so that the anterior margin of the tibial shaft descends steeply from the articular platform. Superior to the tuberosity is a slightly rugose tuberosity bordering the anterior margin of the articular surface, which is separated from the tibial tuberosity by a shallow transverse groove. This is bordered by a faint line medially and a stronger crest laterally, marking the extent of the attachment of the capsule. Lateral to the capsular attachment, and immediately anterior and inferior to the lateral condyle, is a raised, elliptical and slightly convex area that marks the site of the iliotibial tract. On the medial side of the tibial tuberosity the anterior face of the shaft is convex and largely featureless. Numerous vascular foramina are located superiorly, just below to the medial condyle. On the medial side of the shaft, just where the break occurs, there is a small, roughened tuberosity, presumably for the attachment of the tibial collateral ligament. Unfortunately, the site of attachment for the pes anserinus (i.e., the insertion for the sartorius, gracilis and semitendinosus) is not preserved. On the lateral side of the proximal tibia is a roughened area, which arcs postero-superiorly to antero-inferiorly, for the attachments of the tendon of the biceps femoris and the extensor digitorum longus. Inferior to this area, on the lateral side of the tibial tuberosity is a shallow groove, with a maximum width of 8.6 mm, running proximodistally along the shaft. This is the site of attachment for tibialis anterior. A low rounded crest extending down the middle of the lateral side represents the interosseus border. Posterior to this is the area of attachment for the tibialis posterior. The area for tibialis posterior on the lateral side of the shaft is subequal to that of tibialis anterior. Just below the posteromedial lip of the medial condyle is a broad and rugose groove, bordered inferiorly by an irregular crest, which terminates medially in a shallow circular depression. The latter marks the insertion for semimembranosus, which appears to have been well developed. The proximal fibular facet is lanceolate, with a long-axis directed antero-laterally. It is relatively large, measuring 16.3 mm by 11.1 mm. It is well-defined, slightly convex anteroposteriorly and mostly flat mediolaterally. The facet faces distally, posteriorly, and slightly laterally. The fibular facet is located 10.1 mm from the posterolateral margin of the lateral condyle. On the posterior surface, immediately medial to the fibular facet is a deep oval depression that extends a short way down the shaft. It is bordered by a raised, convex area
7 Hominins from Laetoli
medially for the insertion of popliteus, and by a low rounded and ill-defined crest laterally that descends from the fibular facet on the lateral side to at least mid-shaft. The latter crest delimits the posterior border of tibialis posterior and represents the proximal extent of the soleus line.
Comparisons and Functional Implications Compared with modern humans, EP 1000/98 differs in the following respects: the proximal articular surfaces are anteroposteriorly slightly shorter relative to the breadth (with a breadth-length index of 65.5 in EP 1000/98 compared to a mean value in modern humans of 68.0; see Table 7.12); the articular surface of the medial condyle is more concave; the lateral condyle is slightly convex (generally relatively flat to slightly concave in modern humans, although some humans exhibit a slight degree of convexity) (Tardieu 1982, 1983; Berger and Tobias 1996; Organ and Ward 2006); the intercondylar eminence is more elevated, with a stronger transverse crest linking the medial and lateral tubercles; the tibial tuberosity is much less protuberant; the groove superior to the tibial tuberosity is oblique, rather than transversely oriented; the area for the iliotibial tract has greater relief; the area on the lateral side of the tibial tuberosity for attachment of the tibialis anterior is relatively less extensive, and the interosseus line is more anteriorly located; the proximal fibular facet is relatively larger; the area of attachment for the semimembranosus is better developed and forms a distinct circular depression on the posteromedial margin of the condyle; the depression on the posterior aspect of the shaft medial to the fibular facet is deeper; the lateral meniscus has a single tibial insertion in the anterior intercondylar area (humans are unique among primates in having a double insertion, one for each of the anterior and posterior intercondylar areas respectively) (Senut and Tardieu 1985; Tardieu 1986a, b, 1999; Le Minor
171
1990; but see Dugan and Holliday 2009); the cross-section of the shaft at the level of the inferior margin of the tibial tuberosity is less compressed mediolaterally (the breadth-length of the cross-section of the shaft at this level is 95.3 in EP 1000/98, compared with a mean value in modern humans of 75.1; see Table 7.12); and the lateral side of the shaft below the lateral condyle is more deeply hollowed. In addition, EP 1000/98 possesses a weak crest that passes obliquely across the posterior aspect of the shaft, which corresponds to the soleal line in humans. A similar crest is commonly found in the tibiae of chimpanzees, although according to Aiello and Dean (1990) this line does not correspond to the soleal line of humans, because the soleus is normally confined to the fibula in African apes and the crest merely represents the posterior border of the tibialis posterior. However, Gregory (1950) demonstrated that soleus does attach to this crest in Gorilla. Given that a similar crest occurs in humans and chimpanzees, it is not possible to determine whether or not there is a soleus attachment in EP 1000/98 or in other fossil tibiae. Compared with chimpanzees, EP 1000/98 differs in having: a less convex lateral condyle; a relatively more elevated intercondylar eminence; a much less retroverted tibial plateau (in chimpanzees, the average posterior tilt of the medial and lateral condyles relative to the proximodistal long axis of the proximal end of the shaft is 22° and 10° respectively, compared with 4° and 1° in EP 1000/98; see Table 7.12), with less overhang of the condyles posteriorly; the medial and lateral condyles have a similar orientation in the mediolateral plane, whereas in chimpanzees the lateral condyle is more obliquely oriented relative to the transverse plane of the medial condyle; the condyles are proximodistally relatively thicker; the tibial tuberosity is much less protuberant; the area for the iliotibial tract is larger; the proximal fibular facet is relatively smaller; the interosseus line on the lateral side of the shaft is more posteriorly placed;
Table 7.12 Comparison of dimensions (mm) of proximal tibia from Laetoli (EP 1000/98) A. africanusa A. afarensis A. anamensis EP KNM-KP 1000/98 Stw 514a AL 288–1aq AL 129–1b 29285A Dimensionsa
Pan troglodytes
Homo sapiens
N
N
Mean
Range
Mean
Range
AP l prox tibia 33.8 – 32.9 33.3 49.7 10 40.8 39.3–42.6 16 49.4 41.6–53.7 ML br prox tibia 51.6 52.3 50.8 50.6 67.5 10 60.0 55.2–66.2 16 72.6 63.3–80.2 AP l med condyle 33.0 – 31.6 – 43 (e) 10 38.0 36.0–42.6 16 45.6 36.2–51.1 ML br med condyle 20.6 20.5 20.6 – 30 (e) 10 26.6 25.3–28.2 16 31.1 26.0–38.3 AP l lat condyle 26.6 20 (e) 24.8 27.4 – 10 32.6 29.3–37.5 15 38.5 29.2–44.0 ML br lat condyle 22.1 16.8 20.7 22.4 – 10 26.0 23.8–29.3 15 30.7 25.8–33.9 Post slope med condyle 4° – 9° 14° 17° 10 22° 5–30° 14 9° 3–20° Post slope lat condyle 1° – 9° 10° 7° 10 10° 3–20° 14 9° 0–16° AP l shaft 21.2(–) – 27.2 28.9 36.2 10 32.4 28.5–35.2 17 38.2 34.8–41.3 ML br shaft 20.2 – 22.3 22.3 29.0 10 21.6 19.6–24.2 17 28.7 24.2–35.2 AP anteroposterior, br breadth, lat lateral, l length, med medial, ML mediolateral, post posterior, prox proximal, (–) minimum dimension, (e) estimated dimension a Data from Berger and Tobias (1996)
172
the insertion for semimembranosus is larger and better defined; and the proximal end of the shaft is anteroposteriorly relatively shorter and lacks the marked anterior curvature typical of African apes. However, EP 1000/98 does share a number of features with the tibiae of chimpanzees that distinguish them from those of modern humans. These include: a more convex lateral condyle in the anteroposterior plane; a more elevated crest linking the tubercles of the intercondylar eminence; a relatively small and triangular posterior intercondylar region, presumably lacking a posterior attachment for the lateral meniscus; a less protuberant tibial tuberosity; an oblique groove superior to the tibial tuberosity; and a more pronounced hollowing of the lateral aspect of the shaft below the lateral condyle. The features that EP 1000/98 shares with modern humans compared with those of African apes, such as the less pronounced retroversion of the proximal tibial plateau (the posterior tilt of the medial and lateral condyles in the fossil falls at the lower end of the range for modern humans; see Hashemi et al. 2008), a more prominent iliotibial tract for insertion of the tensor fascia latae and gluteus maximus, and a more posterior attachment of tibialis posterior, indicate more extended positions of the knee associated with bipedal locomotion (Aiello and Dean 1990; Berger and Tobias 1996). In humans, individuals with a steeper posterior slope of the tibial plateau is associated with an increase in the magnitude of the anteriorly directed component of the compressive forces that act on the tibial articular surface during the intial phase of weight support (Hashemi et al. 2008). This produces greater anterior translation of the tibia and a higher force exerted by the anterior cruciate ligament. The functional shift to fully extended knees at heel strike in hominins, compared with the weight-bearing semiflexed knees in great apes, would favor the development of a less posteriorly tilted tibial plateau. Nevertheless, a number of primitive features in EP 1000/98 indicate that the knee joint was capable of a greater range of rotation than seen in those of modern humans. These include a more convex lateral condyle in the anteroposterior plane, a relatively larger facet for the proximal fibula, and a single tibial attachment for the lateral meniscus. The hollowing of the lateral side of the shaft below the lateral condyle may imply that the knee was less effective than the modern human knee joint in dissipating high peak loads when the foot made contact with the ground. The relatively welldeveloped insertion for semimembranosus compared with modern humans, as well as the prominent development of the iliotibial tract, implies well-developed capabilities for extension of the thigh at the hip joint. Comparisons with the proximal tibiae of other fossil hominins show that EP 1000/98 is very similar to those of A. afarensis from Hadar. It is comparable in size and closest in overall morphology to the small tibiae from Hadar, such as
T. Harrison
A.L. 288-1aq and A.L. 129-1b, which presumably belonged to female individuals (Johanson and Taieb 1976; Johanson and Coppens 1976; Lovejoy et al. 1982; see Table 7.12). The larger proximal tibiae from Hadar (i.e., A.L. 333x-26 and A.L. 333-42) have linear dimensions that are on average ~29% larger than those of EP 1000/98, and they tend to be more robust, but otherwise they exhibit the same suite of distinctive morphological features. Compared with EP 1000/98, AL 288-1aq and AL 129-1b are very slightly smaller in their proximodistal and anteroposterior dimensions, and the shaft is more bilaterally compressed (the breadth-length index of the cross-section of the shaft at the level of the inferior margin of the tibial tuberosity is 77.2 in AL 288-1aq and 82.0 in AL 129-1b, compared with 95.3 in EP 1000/98; see Table 7.12). The different proportions of the shaft in EP 1000/98 are apparently due to the relatively short anteroposterior dimension of the shaft. Just as in EP 1000/98, the lateral side of the shaft has a distinctive hollowing below the lateral condyle. The morphology of the proximal articular surfaces in the Hadar specimens is identical to that in EP 1000/98, with equal-sized condyles, an anteroposteriorly convex lateral condyle (more marked in the Hadar specimens), similar configuration of the anterior intercondylar area, a relatively restricted posterior intercondylar area lacking a posterior cornu for the lateral meniscus, and a similarly developed intercondylar eminence. On the anterior face of the shaft, the tibiae from Hadar and EP 1000/98 share a relatively low tibial tuberosity (even more weakly expressed in EP 1000/98), an oblique groove on the bursal surface superior to the tibial tuberosity, and a raised surface for a well-developed iliotibial tract. Medially, there is a well-developed depression for the insertion of the semimembranosus. The scar for the pes anserinus, which is very well developed in the proximal tibiae from Hadar, is not preserved in EP 1000/98. In the Hadar specimens, this scar forms a deep groove-like roughened area bordering the raised anterior crest of the shaft. This resembles the condition in African apes more than that in humans, where the attachments for sartorius, gracilis and semitendinosus are placed more laterally and leave a less distinct scar. Laterally, the interosseus line extends down the midline of the shaft, and delimits a relative large area for the attachment of tibialis posterior and a distinctly grooved surface for the tibialis anterior. On the posterior surface of the shaft, there is a pronounced depression below the posteromedial margin of the lateral condyle, bordered laterally by a low rounded crest for the proximal end of the soleal line and medially by a smoothly rounded surface for popliteus. The proximal facet for the fibula is only partially preserved in A.L. 288-1aq, but it does appear to have been relatively smaller than in EP 1000/98. The retroversion of the proximal end is more marked in the Hadar proximal tibiae than in EP 1000/98, and their posterior inclination exceeds or falls
7 Hominins from Laetoli
within the upper end of the range of modern humans (Table 7.12). This latter feature is structurally associated with the more pronounced development of the tibial tuberosity and greater overhang of the condyles posteriorly in the Hadar specimens. The differences between the proximal tibiae from Hadar and Laetoli are relatively minor, however, and the overall morphology of EP 1000/98 is remarkably similar to the two small proximal tibiae of Australopithecus afarensis from Hadar. The only known proximal tibia of Australopithecus anamensis (KNM-KP 29285A from Kanapoi) is much larger than EP 1000/98 (the dimensions of the proximal end are ~39% larger on average, being larger in size than the presumed tibiae of male individuals of A. afarensis from Hadar; Table 7.12), but, like those of A. afarensis, its overall morphology corresponds closely to that seen in EP 1000/98 (Ward et al. 2001). Other than size, KNM-KP 29285A differs in the following respects: the medial condyle has a slightly more undulating articular surface, being concave anteriorly and slightly convex posteriorly (whereas EP 1000/98 is anteroposteriorly more planar); the lateral rim of the medial condyle is more elevated; the intercondylar eminence is relatively larger and more robust; the anterior intercondylar area is more deeply excavated; retroversion of the proximal tibia is more pronounced, with a posterior slope of the medial condyle of 17° and of the lateral condyle of 7°, and a more strongly developed overhang of the condyles posteriorly; the tibial tuberosity and the area superior to the tuberosity for attachment of the capsule are much more rugose; the anterior margin of the shaft is more strongly keeled; the shaft is more strongly bilaterally compressed (the breadth-length index of the cross-section of the shaft at the level of the inferior margin of the tibial tuberosity is 80.1, compared with 95.3 in EP 1000/98; see Table 7.12); the area on the medial side of the condyle that marks the attachment for semimembranosus is more strongly marked; the raised area for the iliotibial tract is larger and more pronounced; the interosseus line is slightly more posteriorly placed, and the area for tibialis anterior is more concave; the posterior face has a more pronounced soleal line. The majority of these differences simply relate to the larger size, greater robusticity, and more pronounced muscle markings of the Kanapoi tibia, and are probably a function of differences that can be attributed to sexual dimorphism rather than being of taxonomic or behavioral importance. In terms of key anatomical features and functional differences, the proximal tibia of A. anamensis is basically indistinguishable from EP 1000/98 and those of A. afarensis (Ward et al. 2001). EP 1000/98 is also similar in size and morphology to the small proximal tibia from Sterkfontein, Stw 514a, belonging to Australopithecus africanus (Table 7.12). Although this latter specimen has been described as chimpanzee-like, with
173
possible adaptations for arboreal locomotor behavior (Berger and Tobias 1996), this interpretation may have been somewhat overstated. The combination of ape-like features of this specimen (i.e., anteroposteriorly convex lateral condyle, unnotched posteromedial margin of the lateral condyle associated with the absence of a posterior cornu of the lateral meniscus, hollowing of the lateral side of the shaft below the lateral condyle, oblique groove superior to the tibial tuberosity, insertion for pes anserinus well-developed and represented by a groove on the medial side of the anterior crest, large circular depression for the attachment of semimembranosus; Berger and Tobias 1996) are typically found in other Pliocene hominins, including Australopithecus anamensis and A. afarensis, as well as EP 1000/98, which all share with modern humans a number of important features of the tibia that are functionally associated with bipedalism. As in EP 1000/98 and other Pliocene hominin tibiae, Stw 514a exhibits a limited degree of retroversion of the tibial plateau, an important feature that it shares with humans in contrast to African great apes. Finally, a poorly preserved proximal tibia from the Upper Burgi Member (~1.9 Ma) is associated with a partial skeleton (KNM-ER 1500) that includes a mandibular fragment that allows attribution of the specimen to Paranthropus boisei (Day et al. 1976; Grausz et al. 1988; Brown and Feibel 1988). The size of the mandible and postcranial elements indicates that the individual was a relatively small female (Grausz et al. 1988). The proximal end of the tibia appears to be slightly larger in overall size than EP 1000/98, although precise measurements are not possible. Although badly weathered and eroded, the anatomical features that are preserved indicate that it was comparable in morphology to EP 1000/98. The key features can be listed as follows: the lateral condyle is convex anteroposteriorly, the posterior intercondylar area is relatively small and lacks a notched posteromedial border of the lateral condyle indicating that there was no posterior cornu for the lateral meniscus, posterior retroversion of the tibial plateau is moderate (about 11° for both the medial and lateral condyles), the remnant of the protuberance for the iliotibial tract is large and well-developed, the site of attachment for the semimembranosus is represented by a large circular depression, the lateral side of the shaft is hollowed below the lateral condyle, and the soleal line is prominent (Day et al. 1976). Few differences separate KNM-ER 1500 from EP 1000/98, with the former having a more retroverted tibial plateau, a more strongly protuberant tibial tuberosity, greater posterior overhang of the medial and lateral condyles (these three features are structurally linked), and a more bilaterally compressed and triangular shaft cross section. KNM-ER 1500 shows the same unique combination of features that are typically found in Australopithecus spp., as well as in EP 1000/98.
174
Comparisons with modern hominoids show that EP 1000/98 can be distinguished from both African great apes and humans. It exhibits a unique combination of features, representing a mosaic of derived specializations associated with bipedalism and a suite of presumably primitive features found in African apes that imply that the knee was capable of a wider range of axial rotation, probably associated with a greater capacity for arboreal climbing. As has been inferred for A. afarensis (Senut 1980; Stern and Susman 1983, 1991; Jungers 1982, 1991; Jungers and Stern 1983; Susman et al. 1984; Tardieu 1986a, b, 1999; McHenry 1991, 1994; Duncan et al. 1994; Stern 2000; Ward 2002), the species to which EP 1000/98 belonged was likely to have been an obligate biped on the ground, but quite effective at climbing in trees to exploit arboreal resources. Comparisons of EP 1000/98 with proximal tibiae belonging to Australopithecus afarensis, A. anamensis, A. africanus and Paranthropus boisei show that they are all morphological similar, with the same suite of features distinguishing them from extant hominoids. These include: medial and lateral condyles of subequal size; anteroposteriorly relatively short condyles; a lateral condyle that is anteroposteriorly convex; an elevated intercondylar eminence with a well-developed crest linking the medial and lateral tubercles; a small and triangular posterior intercondylar area that lacks a posterior attachment for the lateral meniscus; a relatively low degree of retroversion of the tibial plateau; lateral aspect of the proximal shaft that is hollowed below the lateral condyle; an oblique groove superior to the tibial tuberosity associated with the attachment of the capsule; a distinct scar for the pes anserinus on the medial side of the anterior crest (not preserved in EP 1000/98); a well-developed depression on the medial side of the condyle for the semimembranosus; an interosseus line that divides the lateral side of the shaft into subequal areas for the anterior tibialis and posterior tibialis; a pronounced area for the iliotibial tract; possibly a distinct soleal line (but see comments above); and a shaft that is relatively straight anteroposteriorly, without a pronounced anterior curvature. The minor differences between the fossil proximal tibiae from the Pliocene of Africa are far outweighed by the striking uniformity in the key features that they share, and the anatomical differences certainly do not translate into any apparent functional differences. Without a direct association of cranio-dental remains with the proximal tibia from the Upper Ndolanya Beds it is not possible to definitively establish the taxonomic identity of EP 1000/98. The fact that EP 1000/98 resembles the tibiae of other Pliocene hominins from East and South Africa means that there are no distinctive features of the proximal tibia that can be used to discriminate early hominin taxa. Potential hominin species that are broadly contemporary
T. Harrison
with EP 1000/98 include Paranthropus aethiopicus, Australopithecus garhi, Australopithecus africanus, and the earliest representatives of Homo. Unfortunately, however, definitive examples of the proximal tibiae of these taxa are unknown, with the exception of A. africanus. Given that Paranthropus aethiopicus is the only hominin so far represented in the Upper Ndolanya Beds, there is a reasonable likelihood that EP 1000/98 belongs to this species. If so, this would be the first postcranial element attributed to P. aethiopicus. The morphology of the proximal tibia is comparable to that of P. boisei later in time, as well as to the earlier and more primitive species of Australopithecus, so it is possible to deduce that EP 1000/98 is consistent in morphology with what one might anticipate in P. aethiopicus. However, given the uncertainties about the taxonomic attribution of EP 1000/98, the specimen is left unassigned as Hominini gen. et sp. indet. The tibia is very similar in morphology to those of Australopithecus spp. and Paranthropus boisei, and implies that the hominin taxon to which it belonged was comparable to A. afarensis in being a terrestrial biped that was adept at arboreal climbing. EP 1000/98 is comparable in size to the tibiae of female individuals of A. afarensis and P. boisei. The regression formula for the lengths of the medial and lateral condyles for hominoid primates (including humans) published by Jungers (1988) provides an estimated body mass of 37.7 kg for EP 1000/98, which falls in the lower end of the estimated ranges for Australopithecus afarensis (30.4–67.7 kg) and Paranthropus boisei (33.0–69.3 kg).
LAET 75-3817 Zygomatic process of the right frontal of an infant. The specimen is associated with a small cranial fragment and an indeterminate bone fragment, but it is uncertain whether these belong to the same individual. The specimen was originally identified as a cercopithecid, but its size and morphology make it much more likely that it belongs to a hominin. It compares favorably with the morphology of juvenile specimens of Australopithecus, including the A. africanus specimen from Taung, and the A. afarensis specimen, A.L. 333-105 from Hadar. LAET 75-3817 is provisionally identified here as cf. Hominini indet.
Location and Stratigraphic Provenance The specimen was discovered by Mary Leakey’s expedition on August 7, 1975, during excavations of the Upper Ndolanya Beds at Loc. 7E. The specimen was recovered from Strip 8.
7 Hominins from Laetoli
Preservation The specimen consists of the zygomatic process of a right frontal preserving a short section of the lateral orbital margin, the lateral orbital plate, the anteriormost portion of the temporal fossa, the lateral recess of the frontal sinus, a portion of the frontal planum and the endocranial surface of the orbital plate of the frontal. The maximum length of the fragment is 26.8 mm and the maximum perpendicular breadth is 15.5 mm. The sutures for the parietal, sphenoid and zygomatic are still patent and unfused, establishing its juvenile status. In addition, the anterior face of the zygomatic process is pitted by tiny vascular canals, typical of juvenile individuals.
Description The lateral orbital rim of the zygomatic process is mediolaterally slightly convex. The minimum mediolateral breadth between the orbital margin and the lateral margin of the zygomatic process is 7.6 mm (compared with 6.1 mm in Taung and 6.2 mm in A.L. 333-105 from Hadar). Inferiorly it preserves the sutural contact for the zygomatic bone. On the anterior surface, the suture runs mediolaterally and slightly superiorly towards the lateral side. At the junction with the lateral marginal ridge, the suture for the zygomatic bone passes inferiorly. The lateral marginal ridge, separating the anterior face from the temporal fossa, forms a sharp crest. A short section of the ridge is damaged for about 3 mm of its length, causing a shallow irregular pit (possibly caused by termites or safari ants; Hill 1987). The anterior face and temporal fossa meet opposite the zygomatic suture at an angle that slightly exceeds 90° (as in the infant Australopithecus specimens from Taung and Hadar). This indicates a shallow anterior temporal fossa and a limited degree of postorbital constriction (unlike in adult and juvenile cercopithecids, in which the two faces meet at an acute angle greater than 45°). The lateral margin is quite well marked and continues superiorly onto the frontal planum. It was presumably continuous with a relatively pronounced temporal line, but its superior extent is damaged. A large triangular flake of bone has been lost between the suture for the sphenoid and the lateral marginal ridge. This exposes a small area of highly pneumatized bone, infilled with tuffaceous sediment, which probably represents the lateral extent of the frontal sinus. Superiorly the anterior face of the frontal bone exhibits a very slight supero-inferior convexity for the weakly developed superciliary ridge. Superolateral to this ridge the frontal plane is slightly concave to form a shallow sulcus, but otherwise it rises steeply above the orbit. The plane of the frontal just superior to the orbit is angled at 44° relative to the plane of the anterior gutter of the temporal fossa (compared with 45° in the Taung
175
infant), whereas in extant adult cercopithecids the frontal is much more receding (greater than 55°). The configuration of the supraorbital region, with a slight superciliary ridge and steep frontal plane, is closely similar to that seen in Taung and A.L. 333-105. Superiorly, the frontal fragment is broken at a fresh fracture running mediolaterally. The maximum thickness of bone at the point of breakage is only 2.8 mm, which is slightly thinner than that in Taung. The medial aspect of the frontal planum exposes a triangular section through the lateral recess of the frontal sinus. It is heavily pneumatized and filled with matrix. Frontal sinuses are present bilaterally, but variably developed in adult individuals of A. afarensis, P. boisei and P. aethiopicus (Leakey and Walker 1988; Kimbel et al. 2004). Inferiorly, there is a short section (~4 mm long) of the lateral margin of the orbit. The orbit had a relatively sharp lateral border. The internal surface of the orbit is smoothly concave and featureless, except for some tiny pinprick vascular canals. Inferiorly, the orbital plate ends at the suture for the zygomatic bone and a short section of the suture for the sphenoid. The size of the orbit cannot be estimated, but based on its lateral contour it would have been comparable in size to that in Taung. Laterally, the anterior margin of the temporal fossa forms a distinct infero-superiorly aligned gutter. This is due to the prominence of the lateral marginal ridge of the orbital rim, but also because the frontal flares strongly towards the posterior break. This implies an inflated neurocranium laterally and a relatively slight degree of postorbital constriction. In this respect, LAET 75-3817 resembles the configuration seen in Taung. The gutter is bordered inferiorly and laterally by sutures for the sphenoid. The endocranial surface is smooth, except for fine vascular foramina and grooves. The surface is infero-superiorly concave, and mediolaterally convex due to low rounded infero-superiorly directed crest. The two associated bone fragments are much more fragmentary and their precise anatomical location cannot be determined. One piece, a subrectangular plate of bone (29.3 ´ 16.2 mm), represents a fragment from the neurocranium, probably a piece of the parietal or frontal bone. The preservation matches that of the frontal fragment, and they are similar in thickness, but there appears to be no point of contact between them to confirm that they belong to the same individual. The endocranial surface is smooth and featureless, except for a number of tiny vascular canals. The external surface has a low degree of convexity, implying a relatively large neurocranium. A short section of suture (6.7 mm long) is preserved along its shortest margin. The other piece of bone, measuring 14.6 ´ 13.1 mm, is too fragmentary to identify anatomically. It is pentagonal in shape. One of the edges preserves a sharp border and the adjacent side preserves a short section of suture.
176
Comparisons The general morphology of the frontal fragment and the development of the sutures clearly indicate that LAET 75-3817 belongs to a juvenile individual of a large primate. In terms of its size it is derived from an individual somewhat larger than an infant of Papio anubis. It appears to be too large to be attributed to any of the fossil cercopithecids from Laetoli, but does match quite well with the size and morphology of juvenile specimens of A. africanus and A. afarensis. The following features indicate assignment to a hominin rather that to a cercopithecid: (1) lateral rim of the orbit that meets the temporal fossa at an angle greater than 90°; (2) a narrow temporal fossa gutter with a laterally widely flaring neurocranium posteriorly; (3) relatively weakly develop superciliary ridges; (4) a relatively steep frontal planum; and (5) presence of a lateral recess of a frontal sinus.
Discussion The discovery of two new specimens of Australopithecus afarensis (EP 162/00 and EP 2400/00) from the Upper Laetolil Beds, as well as the attribution of several previously unassigned specimens (L.H. 29 and LAET 79-5447), has increased the sample size of A. afarensis from Laetoli from 29 to 33 specimens. The new and previously undescribed specimens include two mandibular corpus fragments and two isolated canines. Although the additional specimens are relatively few, they do help to clarify the extent of morphological and metrical variation in the A. afarensis sample from Laetoli, and they provide an opportunity to reassess the taxonomy and evolutionary relationships. All of the hominins recovered from the Upper Laetolil Beds can be attributed to Australopithecus afarensis Johanson, 1978 (see Table 7.1 for nomenclature and syno nymy list). Some authorities prefer to include this species in Praeanthropus, in conjunction with the growing consensus that the genus Australopithecus sensu lato is a paraphyletic taxon (Skelton et al. 1986; Chamberlain and Wood 1987; Wood 1988, 1991; Skelton and McHenry 1992; Strait et al. 1997; Cameron 2003; Strait and Grine 2004; Kimbel et al. 2004, 2006; White et al. 2006). Given that Australopithecus sensu lato represents a classic Hennigian comb, with good morphological support to infer that A. anamensis, A. afarensis and A. africanus are successively more closely related to Homo or Homo + Paranthropus (Strait and Grine 2004; Kimbel et al. 2004, 2006), there may be justification to include these species, at least, in different genera. As I have stated previously:
T. Harrison “[i]n order for the classification to reflect these inferred relationships, A. africanus, as the type species of the genus, could be retained in Australopithecus or be transferred to Homo, while A. afarensis would need to be removed from Australopithecus and subsequently recognized by the prior name Praeanthropus africanus (Weinert 1950). In my view this option may prove to be a necessary and desirable course of action, but I can fully appreciate that the majority of workers might prefer to retain Australopithecus as a paraphyletic clustering of stem species (just as I do for the Proconsuloidea), at least until such time as the relationships of the early hominids have been more firmly established” (Harrison 1993: 355–356).
Even though uncertainties still exist about the precise relationships between the constituent taxa traditionally included in Australopithecus, there can be little doubt that the genus, as currently construed, is paraphyletic. For the systematic formalist there is only one remedy to this problem – to add new generic names. One solution would be to include all hominin species from A. anamensis onwards in the genus Homo. Another is to include afarensis in the genus Praeanthropus and to create a new monospecific genus for A. anamensis, and presumably a separate one also for A. garhi. Either of these solutions would be perfectly justifiable based on purely phylogenetic grounds, and I am sympathetic to the last of these options, but paleoanthropologists have tended to resist such moves based on our appreciation of early hominin paleobiological diversity. There are good reasons for this resistance, and it stems from the tension that exists between systematic formalism and paleobiology. While many (but not all) paleoanthropologists are willing to accept that Paranthropus represents a clade of hominins morphologically and behaviorally distinct enough to be included in a separate genus from Homo and Australopithecus africanus, the differences between A. africanus, A. afarensis, and A. anamensis are of the kind and degree that neontologists would readily accommodate within a single genus. The case to separate A. afarensis from A. anamensis at the genus level would be particularly problematic, considering that they appear to be closely related members of a single lineage (Ward et al. 1999, 2001; Kimbel et al. 2004, 2006; Kimbel and Delezene 2009; Haile-Selassie et al. 2010). Part of the problem is what I have termed “realized phylogenetic history” (Harrison 1993). Multiple genera are required to accommodate the species included in Australopithecus because individual species of Australopithecus are the sister taxa of descendant species that are not included in the same genus (i.e., Homo and Paranthropus). However, during the Pliocene, before the divergence of Homo and Paranthropus, Australopithecus was monophyletic, and its constituent species were similar enough morphologically and behaviorally to be included in a single genus. This means that the constitution and perception of a genus changes through time as its phylogenetic history is realized (see Harrison 1993). Australopithecus would have represented a clade in the early
7 Hominins from Laetoli
Pliocene, but with the divergence of Homo and Paranthropus in the late Pliocene it became paraphyletic. It can be seen that taxonomic utility is dependent upon whether one’s primary focus is phylogeny or paleobiology. Since most paleoanthropologists are equally concerned with both, the tension between these interests creates a strong motivational force against accepting multiple genera for Australopithecus sensu lato. Paleontologists may prefer to include the species of Australopithecus in multiple genera because it has a realized phylogenetic history, whereas neontologists would consider Australopithecus to be the equivalent of a single extant genus because the constituent species can be inferred to have looked and behaved alike on the Pliocene landscape. There is no easy resolution to this problem, and as a consequence the genuslevel nomenclature of Australopithecus sensu lato is likely to remain in a state of flux for the foreseeable future. Given these issues, the species from Laetoli is provisionally (and conservatively) retained here in Australopithecus (see Table 7.1) rather than included in a monospecific genus, Praeanthropus (see Harrison 1993; Strait et al. 1997; Grine et al. 2006). When Australopithecus afarensis was initially described in 1978 (Johanson et al. 1978) the sample of specimens from Laetoli and Hadar were seen as sufficiently close in morphology to be considered conspecific. Although relatively minor differences in the craniodental morphology and metrics were identified between the two samples (Blumberg and Lloyd 1983; White 1985; Kimbel et al. 1985; Cole and Smith 1987), these were considered to be a consequence of sampling (especially due to the relative paucity of specimens from the type locality of Laetoli) and/or a reflection of intraspecific spatio-temporal variation between different populations of the same species. White (1985) argued that some of the morphological and metrical differences between the two samples were due to the skewed representation of the Laetoli sample towards larger-sized individuals, and that if the sample could be enlarged, the range of variation would likely approximate that seen in the much larger Hadar sample. White’s contention has been affirmed by the discovery of EP 162/00 and EP 2400/00, which are metrically similar to small (presumably female) individuals of A. afarensis from Hadar. These new specimens have helped to close the morphological and metrical gap between the Laetoli and Hadar samples, but a number of differences still remain that appear to consistently distinguish the two samples. Following detailed comparisons, and taking into account the range of variation in the two samples, the material from Laetoli can be distinguished from the specimens from Hadar in the following features: (1) the I2 root is medially positioned relative to the lateral margin of the nasal aperture in the Garusi maxilla (as in A. africanus), whereas in the maxillae from Hadar and that of A. anamensis it is placed lateral to the nasal aperture (Kimbel et al. 2006); (2) the upper
177
canine is more bilaterally compressed (the breadth-length index in the Laetoli sample averages 94.6; n = 4, range 90.3– 98.3) compared with 112.2 in the Hadar sample (n = 12, range 104.5–124.0; Kimbel et al. 2004) (see also White 1985), but similar in proportions to the upper canines of A. anamensis (Ward et al. 2001); (3) P3 has a more asymmetrical occlusal outline, with a mesiodistally longer buccal moiety and a slightly higher paracone (White 1985; Kimbel and Delezene 2009); (4) P4 tends to be relatively broader in relation to its mesiodistal length (mean length-breadth index = 68.4, range = 63.8–75.6, n = 4; compared with the Hadar mean value of 72.4, range = 66.4–81.2, n = 11, but the difference is not statistically significant, Student’s t, P = 0.15); (5) M1 and M2 crowns narrow buccally, with a larger and more distally placed protocone, a smaller hypocone, and a tendency to develop a stronger remnant of the lingual cingulum (Carabelli’s trait) on the mesiolingual margin of the protocone (see also White 1985); (6) M3 is relatively smaller in the Laetoli sample (the area [length x breadth] ranges from 136.5 to 158.5 [n = 4] compared with 149.3–241.2 [n = 9] at Hadar; Kimbel et al. 2004), averaging only 85.7% (n = 2, range = 84.1–87.3) of the area of M2 in associated upper teeth, compared with 98.6% at Hadar (n = 7; range = 88.6–108.3), with reduced distal cusps and distal moiety (see also White 1985; Lockwood et al. 2000; Kimbel and Delezene 2009); and (7) P3 is relatively longer in the Laetoli sample compared with that from Hadar (Lockwood et al. 2000; Kimbel et al. 2006). Although there are difficulties taking standard mesiodistal and buccolingual measurement on P3 because of variation in the orientation and shape of the crown (see White 1977), the mesiodistal length and maximum length dimensions of the P3 crowns from Laetoli do appear to be greater than those from Hadar (even with the addition of the small P3 associated with EP 2400/00). The mesiodistal length of the sample from Laetoli has a mean value of 10.9 mm (range = 9.8–12.2, n = 7) compared to the mean value at Hadar of 9.2 mm (range = 7.9– 11.4, n = 19; Kimbel et al. 2004). This difference is statistically significant (Student’s t, P = 0.0001). Maximum length, regardless of orientation of the tooth, provides a simpler measure of crown length, one that is more easily replicated. In this case, L.H. 3 (13.3 mm), L.H. 14 (left = 13.0 mm; right = 13.3 mm) and L.H. 24 (13.1 mm) fall outside the range for the Hadar sample (9.5–13.0 mm; Kimbel et al. 2004), whereas only EP 2400/00 (11.9 mm), L.H. 2 (10.8 mm), and L.H. 4 (12.6 mm) fall within the range. The mean value for the Laetoli sample is 12.6 compared with 11.4 mm for the Hadar sample, and the difference between the two samples is statistically significant (Student’s t, P = 0.0421). By comparison, the mean maximum length of P3 in A. anamensis is 12.2 mm (Ward et al. 2001), which is smaller than in the Laetoli sample, but not significantly different (Student’s t, P = 0.3796).
178
A number of other possible morphological differences have been reported in the literature, but these appear to be variable features that do not consistently differentiate the two samples. These are briefly reviewed below. 1. Configuration of the lateral margin of the nasal aperture: A number of authors (Puech et al. 1986; Leakey et al. 1995; Ward et al. 2001; Kimbel et al. 2004, 2006) have noted that the Garusi maxilla from Laetoli differs from all of the specimens from Hadar in that the facial aspect of the maxilla grades smoothly into the inferolateral margin of the nasal aperture. In the Hadar maxillae, by contrast, a sharp crest delimits the inferolateral margin of the nasal aperture. In this respect, the Garusi maxilla is similar to the maxilla of A. anamensis (KNM-KP 29283) from Kanapoi (Leakey et al. 1995; Ward et al. 2001; Kimbel et al. 2004, 2006). Further comparisons by the author have confirmed the validity of this difference, but unfortunately we do not know enough about the variability of this feature in early hominin populations to assess its taxonomic significance. For example, we only have one example of an adult maxilla at Laetoli and Kanapoi respectively that preserves this region. However, it is important to note that the right maxillary fragment associated with the partial skeleton of an infant from Laetoli, L.H. 21, has a sharp crest forming the lateral margin of the nasal aperture, just as in the adult maxillae from Hadar. 2. Narrower and shallower palate: Ward et al. (1999) have suggested that the anterior palate of the Garusi specimen from Laetoli resembles A. anamensis in being narrower and shallower than in the Hadar maxillae (see also Puech et al. 1986). However, Kimbel et al. (2004) have shown that the Garusi maxilla is not different from the Hadar maxillae in these respects. The internal palate breadth at P3 in the Garusi maxilla can be estimated to be 26.2 mm, which corresponds well with comparable dimensions in the maxillae from Hadar. 3. Inclination and external contour of the mandibular symphysis: The only mandibular specimen of an adult from Laetoli preserving the symphyseal region is L.H. 4. Kimbel et al. (2004, 2006) have suggested that the Laetoli specimen has a more strongly inclined symphyseal region than in the mandibles from Hadar, and that the external contour of the symphysis is more convex and inferiorly “cut away”, compared to the straighter external surface superiorly with a basal bulge seen in Hadar. However, there is a good deal of variation in the Hadar mandibles. The inclination of the symphysis in L.H. 4 is 127° relative to the alveolar plane of the tooth row, which falls in the upper end of the range of the sample from Hadar (mean = 119°, range 105–134°, n = 11), and can be matched or exceeded by several Hadar specimens with more inclined symphyses (i.e., A.L. 333w-12, A.L. 198-
T. Harrison
1, A.L. 266-1). Comparison of the contour of the external face of the symphysis is more difficult to assess objectively, but, while many of the Hadar mandibles exhibit a steeply inclined superior symphyseal region with a bulging basal aspect, a number of them (i.e., A.L. 207-13, A.L. 315-22, A.L. 330-5 and A.L. 333w-60, A.L. 438-1 g; see Kimbel et al. 2004, fig. 5.29) closely resemble the configuration seen in L.H. 4. Although the symphyseal region of L.H. 4 is not typical of that seen at Hadar, its morphology is certainly encompassed by the range of variation. 4. Position of the lower canine in relation to the postcanine toothrow: The larger lower canine roots in A. anamensis results in a more laterally expanded canine jugum compared with that in A. afarensis. As a result, the mandible has its greatest anterior breadth at the level of the canine in A. anamensis, whereas in A. afarensis it is greatest at P3 (Ward et al. 2001). In other words, the lateral aspect of the mandibular corpus curves medially anterior to the lower canine in A. anamensis, whereas it curves medially anterior to P3 in A. afarensis from Hadar. Kimbel et al. (2006) contend that the morphology in L.H. 4 is intermediate, with the canine in line with the postcanine long-axis and a medial curvature at C/P3. However, the P3 jugum in L.H. 4 is more pronounced than in A. anamensis and its configuration closely resembles that in the Hadar sample. A way to quantify the difference between A. anamensis and A. afarensis in canine position is to measure by how much the long axis of the lower canine deviates medially or laterally from the long axis of the postcanine tooth row. In L.H. 4 the canine deviates medially by a distance of 3.2 mm, which falls within the range of variation of mandibles from Hadar (1.9–4.6 medial deviation, n = 4), whereas in A. anamensis (KNM-KP 29281A) the canine long-axis deviates laterally by 1.0 mm. Although the canine is missing in EP 2400/00, the posterior half of the canine alveolus is preserved, and it is evident that the canine was placed medial to the long-axis of the postcanine tooth row, the P3 jugum was more pronounced than the canine jugum, and the lateral aspect of the mandible curved medially just anterior to P3, as in the mandibles from Hadar. In this regard, the Laetoli mandibles do not differ from those from Hadar. 5. Symmetry of the mesial and distal crown shoulders: Kimbel et al. (2006) argue that the Hadar upper canines differ from those from Laetoli in having a more asymmetrical crown in lingual view because the mesial shoulder is more apically placed than the distal shoulder, whereas in the upper canines from Laetoli the shoulders are similar in elevation as in A. anamensis (Ward et al. 2001). This distinction is valid when the Hadar upper canines are compared with L.H. 3, which appears to be unusual in having mesial and distal shoulders at the same elevation. However, L.H. 5, L.H. 6 and LAET 79-5447
7 Hominins from Laetoli
(see Fig. 7.8) are more similar to A.L. 333x-3 in their degree of asymmetry, with a more apically placed mesial shoulder. Besides, the best-preserved upper canine of A. anamensis (KNM-KP 35839) (Ward et al. 1999, 2001), which is considered to exhibit the morphology typical of the species (see Ward et al. 2001: 347), has a suite of peculiar features (i.e., strongly developed and basally placed lingual cingulum, a simple lingual face with no pillar, straight mesial and distal crests, a lingually recurved crown, and a slender root) that leaves me unconvinced that this tooth (and the associated “pathological” incisor) is representative of the taxon. 6. Frequency of occurrence of a prominent metaconid on P3: White (1985) noted that there may be a higher frequency of P3 crowns from Laetoli with a prominent metaconid compared with the sample from Hadar. However, additional data from Laetoli show that the frequency of occurrence of P3s with a small metaconid is similar to that at Hadar (33.3% at Laetoli [n = 6] compared with 40% at Hadar [n = 10]). This contrasts with the more primitive condition in A. anamensis in which 100% (n = 5) of the crowns lack a distinct metaconid (Ward et al. 2001). A case has been made that the Laetoli specimens are morphologically (and temporally) intermediate between A. anamensis and the Hadar samples and that the principal samples of A. anamensis (from Kanapoi and Allia Bay) and A. afarensis (from Laetoli and Hadar) represent a single anagenetically evolving lineage (or evolutionary species) that extends through time by more than one million years (Ward et al. 1999; Lockwood et al. 2000; Kimbel et al. 2006; White et al. 2006). This has naturally led some authors to question whether A. anamensis and A. afarensis should be considered separate species, and if so, where to draw the taxonomic line between them (Kimbel et al. 2006; Haile-Selassie et al. 2010). One solution would be to include all of the samples in a single species, A. afarensis (Senut 1996; Wolpoff 1999; Kimbel et al. 2006; Haile-Selassie et al. 2010). If one prefers a two-species model, however, the following taxonomic options are conceivable: (1) [Kanapoi = A. anamensis] [Allia Bay + Laetoli + Hadar = A. afarensis]; (2) [Kanapoi + Allia Bay = A. anamensis] [Hadar + Laetoli = A. afarensis] – currently the consensus view; and (3) [Kanapoi + Allia Bay + Laetoli = A. afarensis] [Hadar = A. antiquus] (see also Kimbel et al. 2006). However, when the evidence is critically scrutinized an entirely different perspective on this problem emerges. While it is justified to conclude that the Laetoli specimens retain a few primitive characteristics that they share with A. anamensis not present in the material from Hadar, it is not the case that the Laetoli sample is intermediate in morphology between the samples from Kanapoi + Allia Bay and Hadar when judged in the broader context of their overall morphology. As noted
179
above, there are relatively few features that consistently differentiate the Laetoli and Hadar samples, whereas the morphological divide between A. afarensis (Laetoli + Hadar combined) and A. anamensis is much more profound. Of the features identified above that consistently differentiate the Laetoli and Hadar samples, only the greater bilateral compression of the upper canines, the relatively longer P3, the higher frequency of remnants of the lingual cingulum on the upper molars, the asymmetry of the P3, and possibly the relatively broader P4 are more primitive features that the Laetoli sample shares with A. anamensis (Ward et al. 1999, 2001; Kimbel et al. 2006; White et al. 2006). On the other hand, the position of the I2 root relative to the lateral margin of the nasal aperture, the higher frequency of a prominent metaconid on P3, the shape and occlusal morphology of the upper molars, and the relatively small size of M3 in the Laetoli specimens are either autapomorphies or are more derived features shared with A. africanus (in the case of the I2 root position and the higher frequency of well-developed P3 metaconids). They cannot be inferred to be intermediate between A. anamensis and the Hadar sample. In contrast, the differences that distinguish A. anamensis from A. afarensis are much more extensive. These include: articular eminence less well-developed; smaller elliptical external acoustic meatus; larger upper canine roots and more prominent canine juga; palate relatively narrower, with tooth rows almost parallel rather than posteriorly diverging; more strongly inclined mandibular symphysis; longer postincisive planum; mandibular corpus less robust below the molars; lower postcanine tooth rows set close together and more parallel, rather than posteriorly diverging; lower canines more laterally placed relative to the postcanine toothrow, with more prominent juga (in A. afarensis the lateral aspect of the corpus turns medially just anterior to P3); upper canine with a stronger mesiolingual ridge; I2 is mesiodistally relatively broader; lower canine with pronounced distal heel and basal cingulum; P3 consistently unicuspid with the metaconid represented by a tiny tubercle on the distolingual crest (in A. afarensis more than half of the specimens have a prominent metaconid); mesial fovea of P3 opens lingually with a notched mesial marginal ridge (i.e., lingual cingulum); P4 with less expanded distal fovea; molars are lowercrowned, with greater buccolingual flare; lower molars with a higher frequency of a prominent vestige of the buccal cingulum (protostylid) (40% of A. anamensis lower molars, compared with only 6.5% in the combined Hadar-Laetoli sample; Hlusko 2004); dP3 (=dm1) narrower, with a relatively weakly developed mesial fovea and talonid basin; capitate with laterally facing metacarpal II facet (Leakey et al. 1995, 1998; Ward et al. 1999, 2001; Kimbel et al. 2006; White et al. 2006). The main point to underscore here is that the morphological and metrical differences between the Laetoli and Hadar
180
T. Harrison
Fig. 7.11 Alternative views of the relationship between the A. afarensis samples from Laetoli (L) and Hadar (H), and the A. anamensis samples from Kanapoi (K) and Allia Bay (A). The model on the left, favored by Kimbel et al. (2006), explains the relationship between A. anamensis and A. afarensis as the result of anagenesis in a single evolutionary lineage,
with each step being separated by small morphological changes through time. The model on the right, preferred by the present author, explains the relationship as a cladogenetic event, with minor progressive transformations occurring within the two lineages, but a morphological divide separating A. anamensis and A. afarensis into distinct sister species
samples are few in number and constitute relatively minor distinctions. Such differences in dental morphology are what one might expect as part of the normal variation seen in populations of a single species of hominoid. Rather than being intermediate in morphology, the evidence indicates that the Laetoli sample represents an earlier population of A. afarensis, with almost the full complement of derived features that characterizes the Hadar sample, but, consistent with its greater antiquity, still retaining a small number of more primitive traits. By comparison, the suite of morphological features that distinguishes A. anamensis (at least from the type locality of Kanapoi) from A. afarensis from Laetoli and Hadar is much more extensive and more substantial in nature, and these clearly provide adequate grounds for the recognition of a species distinction. It is also worth emphasizing here that the temporal gap between the youngest sample of A. anamensis from Allia Bay (~3.9 Ma) and the oldest specimens of A. afarensis from Laetoli (3.81–3.83 Ma) is probably as little as 70–90 kyrs. A reasonably good case can be made to document temporal trends within both A. afarensis and A. anamensis and to infer an ancestral-descendant relationship between A. anamensis and A. afarensis (Ward et al. 1999, 2001; Lockwood et al. 2000; Kimbel et al. 2004, 2006), but the current evidence favors an evolutionary model
involving a cladogenetic event rather than a simple anagenetic transformation of a single unbranched lineage through time (see Fig. 7.11). The recent description of a sample of Australopithecus sp. from Woranso-Mille in Ethiopia, contemporaneous (3.57– 3.8 Ma) with the sample from Laetoli, adds another level of complexity to unraveling the relationships between A. anamensis and A. afarensis, by showing that contemporary populations have novel mosaics of cranio-dental features (Haile-Selassie et al. 2010). The material demonstrates how contemporaneous populations of hominins are likely to be characterized by spatial heterogeneity and that one cannot expect that all populations will conform uniformly to simple models of progressive transformation through time. For example, the Woranso-Mille specimens exhibit a less inclined mandibular symphysis and P3 with a prominent metaconid and a well-developed mesial marginal ridge, features that are typical of A. afarensis at Hadar and Laetoli. However, in several features the Woranso-Mille specimens resemble A. anamensis. These include greatest anterior breadth of the mandible at the canine rather than P3 (although the accompanying illustration in Haile-Selassie et al. (2010) appears to show the canine positioned medial to the postcanine tooth row as in A. afarensis) and P4 with a small and elevated trigon basin. Based on the
7 Hominins from Laetoli
descriptions presented by Haile-Selassie et al. (2010) the Woranso-Mille material seems to be very close in morphology to the contemporary sample from Laetoli, but differs in having a relatively larger M3. In my view the material is best attributed to A. afarensis, but Haile-Selassie et al. (2010) remain undecided about its taxonomic status, so it is best to await more detailed comparisons (and perhaps more material) before making a definitive assignment. In addition to new specimens from the Upper Laetolil Beds, this report describes the first hominins from the Upper Ndolanya Beds. These are derived from deposits dated to 2.66 Ma (Deino 2011), more than one million years younger than the A. afarensis specimens from the Upper Laetolil Beds. Two specimens have been recovered since 1998, a maxilla and a proximal tibia. The maxilla (EP 1500/01) can be definitively attributed to Paranthropus aethiopicus, based on its unique combination of features shared with KNM-WT 17000 from the Nachukui Formation, West Turkana, in northern Kenya (Walker et al. 1986; Leakey and Walker 1988; Walker and Leakey 1988; Kimbel et al. 1988; Wood and Constantino 2007). Taxonomic attribution of the tibia is uncertain, because it is not possible to distinguish isolated postcranial elements of Pliocene hominins. Apart from being the first records of hominins from the Upper Ndolanya Beds, these finds are important for several other reasons: (1) the maxilla represents the first definitive record of P. aethiopicus outside the Turkana Basin of northern Kenya and southern Ethiopia; (2) it is among the oldest known records of this species; and (3) the proximal tibia is the first postcranial element of an adult individual to be recovered from Laetoli. The broader importance of these specimens is briefly discussed below. Previously, Paranthropus aethiopicus was known only from the Omo Shungura Formation, Ethiopia (including the holotype, Omo 18–18) and the Nachukui Formation, Kenya, on the northern and western side of the Turkana Basin respectively. The new specimens from the Upper Ndolanya Beds thus extend the range of the species from the Turkana Basin to northern Tanzania, almost 800 km to the south. A maxillary specimen of Paranthropus from the Pliocene Chiwondo Beds at Malema, Malawi, a locality situated 800 km to the south of Laetoli, could potential represent another specimen of P. aethiopicus outside the Turkana Basin (Kullmer et al. 1999). The associated fauna has been estimated to be 2.3– 2.5 Ma, which overlaps chronologically with the known time range of P. aethiopicus (~2.3–2.7 Ma), although it could be considerably younger (Hill 1999). Nevertheless, the morphology of the specimen indicates that it is best attributed to P. boisei rather than P. aethiopicus (Kullmer et al. 1999). Most of the specimens attributed to P. aethiopicus from the Turkana Basin, including the cranium KNM-WT 17000, range in age from 2.33 to 2.58 Ma, but a few specimens from the Omo Shungura Formation (i.e., L55-s-33, L 62-17, Omo
181
18–18, and Omo 84–100) occur in earlier horizons dating to between 2.58 and 2.7 Ma (Feibel et al. 1989; Suwa 1988; Suwa et al. 1996; Wood and Constantino 2007). Omo 18–18 and Omo 84–100 are close in age to 2.6 Ma, while the mandibular fragment L55-s-33 and the isolated molar L 62-17 from Submembers C-6 and C-5 of the Omo Shungura Formation respectively have an estimated age of ~2.70 Ma (Feibel et al. 1989; Suwa et al. 1996; Bobe and Leakey 2009). These latter specimens are attributable to Paranthropus, and can be assumed to be P. aethiopicus, although they do not preserve any diagnostic features (Suwa et al. 1996). The maxilla from the Upper Ndolanya Beds, dating to 2.66 Ma, is thus the oldest, securely dated specimen definitively attributable to P. aethiopicus. The occurrence of P. aethiopicus in the Upper Ndolanya Beds has important implications for understanding and interpreting the biogeography of Paranthropus. The evidence suggests that soon after its earliest appearance in East Africa, at about 2.7 Ma, or shortly thereafter, it attained a relatively wide distribution throughout the region occurring from southern Ethiopia to northern Tanzania (Wood and Constantino 2007; Constantino and Wood 2007). Since there is no immediate precursor for P. aethiopicus in the Turkana Basin or elsewhere in eastern Africa prior to 2.7 Ma (but see Rak et al. 2007), it might imply that the Paranthropus clade originated outside of the geographical province and immigrated into the region. Suwa et al. (1996), however, have tentatively identified a few isolated teeth of relatively large size that might document an earlier occurrence of the Paranthropus clade in the Omo Shungura Formation, dating back to ~2.9 Ma. This would push back the age of Paranthropus in eastern Africa, but it does not change the overall biogeographic implications. The earlier species, A. afarensis, that occupied much of the region until at least 3.0 Ma, is probably morphologically too distinct from Paranthropus aethiopicus to have been its immediate ancestor (see Kimbel and Delezene 2009; but see Rak et al. 2007). The same conclusion might be reached for early Homo (Pickford 2004), which also makes an appearance in eastern Africa in the mid-Pliocene (Kimbel et al. 1996; Kimbel, 2009). However, in this case, differences in the timing and distribution might imply that Homo and Paranthropus do not have coincident biogeographic, ecological and immigration histories. The earliest occurrences of Homo are from Member E of the Omo Shungura Formation (2.4 Ma), the Kalochoro Member of the Nachukui Formation (2.3 Ma), the Busidima Formation at Hadar (2.33 Ma), the Chemeron Formation (~2.4 Ma) and the Chiwondo Beds at Uraha (~2.3–2.5 Ma?) (see Kimbel 2009). Contemporaneous with these occurrences and slightly older are archaeological sites with Oldowan stone tools, which extend back to 2.6–2.5 Ma in the Gona region of the Upper Awash (Semaw et al. 1997, 2003).
182
It is interesting, in this regard, that the Upper Ndolanya Beds at 2.66 Ma only record the presence of Paranthropus aethiopicus, with no evidence of early Homo or archaeological traces. Lithic artifacts have been reported as surface finds from the Upper Ndolanya Beds at Loc. 18 (Kaiser et al. 1995), but no similar finds have been located in situ despite extensive investigations in these horizons by successive teams working at Laetoli, and I am inclined to discount these artifacts as being intrusive from the overlying Ngaloba Beds. In this case, the evidence indicates that the arrival of Paranthropus and Homo in different parts of eastern Africa was not synchronous, and that at Laetoli Paranthropus colonized the area ahead of Homo. A similar pattern is seen in the Turkana Basin, where the paleontological and archaeological record is much more complete. The evidence indicates that Paranthropus was present in the Turkana Basin by at least 2.7 Ma, whereas the earliest record of Homo and of archaeological sites is at 2.4 Ma (Kibunjia et al. 1992; Prat et al. 2005; Roche et al. 2009; Bobe and Leakey 2009). However, at the other end of the Rift Valley, north of the Turkana Basin, in the Awash region of Ethiopia, archaeological evidence implies that early Homo was present by 2.6 Ma, possibly slightly earlier than in the East African Rift to the south, whereas the first record of Paranthropus is from Konso at 1.4 Ma. Homo and Paranthropus may have colonized eastern Africa from different geographic origins, with Homo first appearing at the northern end of the Rift Valley and Paranthropus first appearing to the south. Detailed comparisons of the faunal, ecological and biogeographical relationships between localities in eastern Africa might yield some valuable clues to understanding the nature and timing of the appearance of Paranthropus and Homo in the fossil record.
Conclusions Renewed investigations at Laetoli and at other sites on the Eyasi Plateau since 1998 have led to the recovery of additional fossil hominins. Two specimens, an isolated lower canine (EP 162/00) and a mandibular fragment with P3–M1 (EP 2400/00), recovered from the Upper Laetolil Beds at Loc. 16, are referable to A. afarensis. In addition, two hominins, a proximal tibia (EP 1000/98) and an edentulous maxilla (EP 1500/01), were recovered from the Upper Ndolanya Beds, and represent the first hominins from this stratigraphic unit. EP 1500/01 represents the only specimen of P. aethiopicus recovered from outside the Turkana Basin. A detailed descriptive account of the morphology of these newly collected hominins from Laetoli is presented above. In addition, four specimens from the Kohl-Larsen and Mary Leakey collections are described in detail for the first time. Three of these specimens, L.H. 29, LAET 79-5447 and MB Ma. 8294,
T. Harrison
are of unknown stratigraphic provenance, but they are almost certainly derived from the Upper Laetolil Beds. All are attributable to A. afarensis. Additionally, a specimen (LAET 75-5447) from the Upper Ndolanya Beds collected by Mary Leakey, and originally identified as a cercopithecid, most probably represents a cranial fragment of a hominin infant. Renewed study of the hominins from Laetoli also provided an opportunity to clarify their chronology and provenance. Fossil hominins have not yet been recovered from the Lower Laetolil Beds (~4.4–3.85 Ma), and they are rare in the younger stratigraphic units that overlie the Upper Laetolil Beds. A single cranium of Homo sapiens is known from the Late Pleistocene Upper Ngaloba Beds. The hominins from the Upper Laetolil Beds, dated from 3.63 to 3.85 Ma, are all referable to Australopithecus afarensis. Except for a single partial skeleton of an infant (L.H. 21), the sample (n = 33) consists entirely of cranio-dental specimens (see Su and Harrison 2008). Australopithecus afarensis is replaced in the fossil-rich Upper Ndolanya Beds, dated to 2.66 Ma, by Paranthropus aethiopicus. The majority of specimens from Laetoli have been recovered as surface finds, and it is important to distinguish between the actual find spot and the stratigraphic horizon from which the fossil originated. The precise stratigraphic provenances listed by Leakey (1987a) are actually the find spots, with the exception of those few specimens recovered from in situ. However, the original stratigraphic provenance of most finds can be deduced with a reasonable degree of precision, at least in relation to the designated Upper Laetolil marker tuffs. Given that Australopithecus sensu lato likely represents a paraphyletic group, there may be justification to include the constituent species in multiple genera. For this reason, some authors advocate including afarensis in the genus Praeanthropus. Although this is a perfectly justifiable move based on purely phylogenetic grounds, and one to which I am sympathetic, I provisionally retain afarensis here in Australopithecus to reflect its anatomical and paleobiological proximity to the other species of Australopithecus sensu lato. In my view, the differences between A, africanus, A. afarensis and A. anamensis are of the kind and degree that neontologists would easily accommodate within a single genus. It has been argued that the Laetoli sample of A. afarensis is morphologically (and temporally) intermediate between A. anamensis and the Hadar sample, and that the principal samples of A. anamensis (from Kanapoi and Allia Bay) and A. afarensis (from Laetoli and Hadar) represent a single anagenetically evolving lineage (Ward et al. 1999; Lockwood et al. 2000; Kimbel et al. 2006; White et al. 2006). However, the new specimens from the Upper Laetolil Beds have helped to close the morphological and metrical gap between the Laetoli and Hadar samples of A. afarensis, and a critical assessment of the morphological variation in the two samples
7 Hominins from Laetoli
indicates that there are relatively few consistent differences separating them. The Laetoli specimens retain a few primitive characteristics that they share with A. anamensis not present in the material from Hadar, but the Laetoli sample cannot be considered to be intermediate in morphology. Only a few features consistently differentiate the Laetoli and Hadar samples, and these are the type of differences that one would expect for intraspecific variation in populations separated in space and time, whereas the morphological divide between A. afarensis (Laetoli + Hadar combined) and A. anamensis is much greater. Rather than being intermediate in morphology, the Laetoli sample appears to represents an earlier population of A. afarensis, with almost the full complement of derived features that characterizes the Hadar sample, but, consistent with its greater antiquity, still retaining a small number of more primitive traits. By comparison, the suite of morphological features that distinguishes A. anamensis from A. afarensis is much more extensive and more substantial in nature, and these clearly provide adequate grounds for the recognition of a species distinction. The current evidence favors an evolutionary model involving a cladogenetic event rather than a simple anagenetic transformation of a single unbranched lineage through time. The hominins from the Upper Ndolanya Beds are important because they provide the first record of P. aethiopicus outside the Turkana Basin, and the oldest securely dated specimen definitively attributable to this taxon. The occurrence of P. aethiopicus at Laetoli also has important implications for understanding and interpreting the biogeography of Paranthropus. The evidence suggests that soon after its earliest appearance in East Africa at about 2.7 Ma, it established a relatively wide distribution throughout the region occurring from southern Ethiopia to northern Tanzania. Since there is no immediate precursor for P. aethiopicus in eastern Africa prior to 2.7 Ma, the Paranthropus clade probably originated outside of the geographical province and immigrated into the region. The earlier species, A. afarensis, that occupied much of the region until at least 3.0 Ma, is probably morphologically too distinct from Paranthropus aethiopicus to have been its immediate ancestor (but see Rak et al. 2007). Homo also appears to have been an immigrant into eastern Africa, rather than derived autochthonously from a local ancestral species. However, the timing of its first appearance and its geographic distribution at Pliocene localities suggests that the dispersal events of Homo and Paranthropus were not coincident or synchronous. The evidence from Laetoli indicates that Paranthropus was present locally by at least 2.66 Ma in the absence of any trace, paleontological or archaeological, of early Homo. A similar pattern is found in the Turkana Basin where the earliest record of Paranthropus at ~2.7 Ma precedes that of Homo by about 300 kyrs (Roche et al. 2009; Bobe and Leakey 2009). In contrast, the Awash region of Ethiopia well to the north of the Turkana Basin has
183
archaeological evidence at 2.5–2.6 Ma that implies that early Homo was present much earlier than Paranthropus. These differences in the timing and mode of immigration into eastern Africa suggests that Paranthropus and Homo had different biogeographic histories, and that the ancestral species may have had slightly different ecological requirements at the time of their initial influx into the region. Acknowledgements The author is grateful to the Tanzania Commission for Science and Technology and the Unit of Antiquities in Dar es Salaam for permission to conduct research in Tanzania. Special thanks go to Paul Msemwa (Director) and Amandus Kweka, as well as to all of the staff at the National Museum of Tanzania in Dar es Salaam, for their support and assistance. The Government of Kenya and the National Museums of Kenya are thanked for permission to study the collections in Nairobi. Thanks go to Emma Mbua, Meave Leakey (Kenya National Museum), Peter Andrews, Jerry Hooker, Chris Stringer and the library staff (Natural History Museum, London), Oliver Hampe, Wolf-Dieter Heinrich (Humboldt-Universität Museum für Naturkunde, Berlin), N. Conard, M.N. Haidle (Institut für Ur- und Frühgeschichte und Archäologie des Mittelalters, Tübingen), Nancy Simmons, Ross MacPhee, Eric Delson, and Eileen Westwig (American Museum of Natural History, New York) for access to specimens in their care. W.H. Kimbel, C. Ward and A. Hill reviewed the manuscript and gave thoughtful comments that greatly improved the quality of the end-product. For their advice, discussion, and help I gratefully acknowledge the following individuals: P. Andrews, R. Bobe, E. Delson, P. Ditchfield, C. Harrison, T.S. Harrison, C. Jolly, D.M.K. Kamamba, W.H. Kimbel, A. Kweka, M.G. Leakey, O. Hampe, M.L. Mbago, C.S. Msuya, S. Odunga, M. Pickford, Y. Rak, C. Robinson, D. Su, C. Ward, and B. Wood. Research on the Laetoli hominins was supported by grants from the National Geographic Society, the Leakey Foundation, and NSF (grants BCS-9903434 and BCS-0309513).
References Adloff, P. (1908). Die Zähne des Homo primigenius von Krapina. Anatomischer Anzeiger, 32, 301–302. Agnew, N., & Demas, M. (1998). Preserving the Laetoli footprints. Scientific American, 279, 44–55. Aiello, L., & Dean, C. (1990). An introduction to human evolutionary anatomy. London: Academic. Alemseged, Z., Spoor, F., Kimbel, W. H., Bobe, R., Geraads, D., Reed, D., & Wynn, J. G. (2006). A juvenile early hominin skeleton from Dikika, Ethiopia. Nature, 443, 296–301. Arambourg, C., & Coppens, Y. (1967). Sur la découverte, dans le Pléistocène inférieur de la vallée de l’Omo (Éthiopie), d’une mandibule d’Australopithécien. Comptes Rendus de l’Académie des Sciences, Paris, Série D, 265, 589–590. Arambourg, C., & Coppens, Y. (1968). Découverte d’un Australopithécien nouveau dans les gisements de l’Omo (Éthiopie). South African Journal of Science, 64, 58–59. Asfaw, B., White, T., Lovejoy, O., Latimer, B., Simpson, S., & Suwa, G. (1999). Australopithecus garhi: A new species of early hominid from Ethiopia. Science, 284, 629–635. Berge, C., Penin, X., & Pellé, É. (2006). New interpretation of Laetoli footprints using an experimental approach and Procrustes analysis: Preliminary results. Comptes Rendus Paleovol, 5, 561–569. Berger, L. R., & Tobias, P. V. (1996). A chimpanzee-like tibia from Sterkfontein, South Africa and its implications for the interpretation
184 of bipedalism in Australopithecus africanus. Journal of Human Evolution, 30, 343–348. Blumberg, B., & Lloyd, A. T. (1983). Australopithecus and the origin of the genus Homo: Aspects of biometry and systematics with accompanying catalog of tooth metric data. Biosystems, 16, 127–167. Bobe, R., & Leakey, M. G. (2009). Ecology of Plio-Pleistocene mammals in the Omo-Turkana Basin and the emergence of Homo. In F. E. Grine, J. G. Fleagle, & R. E. Leakey (Eds.), The first humans: Origin and early evolution of the genus Homo (pp. 173–184). Dordrecht: Springer. Bory de Saint-Vincent, J. B. G. M. (1825). Homme, Homo. In J. V. Audouin (Ed.), Dictionnaire Classique d’Histoire Naturelle (Vol. 8, pp. 269–346). Paris: I. Rey and Gravier. Broom, R. (1936). A new fossil anthropoid skull from Sterkfontein, near Krugersdorp, South Africa. Nature, 138, 486–488. Broom, R. (1938). The Pleistocene anthropoid apes of South Africa. Nature, 142, 377–379. Brown, F. H., & Feibel, C. S. (1988). “Robust” hominids and PlioPleistocene paleogeography of the Turkana Basin, Kenya and Ethiopia. In F. E. Grine (Ed.), Evolutionary history of the “Robust” Australopithecines (pp. 325–341). New York: Aldine de Gruyter. Brunet, M., Beauvilain, A., Coppens, Y., Heintz, E., Moutaye, A. H. E., & Pilbeam, D. (1996). Australopithecus bahrelgazali, une nouvelle espèce d’Hominidé ancien de la region de Koro Toro (Tchad). Comptes Rendus de l’Académie des Sciences, 332(Série IIa), 907–913. Cameron, D. W. (2003). Early hominin speciation at the Plio/Pleistocene transition. Homo, 54, 1–28. Campisano, C. J., & Feibel, C. S. (2008). Depositional environments and stratigraphic summary of the Pliocene Hadar Formation at Hadar, Afar depression, Ethiopia. Geological Society of America Special Paper, 446, 179–201. Chamberlain, A., & Wood, B. A. (1987). Early hominid phylogeny. Journal of Human Evolution, 16, 119–133. Charteris, J., Wall, J. C., & Nottrodt, J. W. (1981). Functional reconstruction of gait from the Pliocene hominid footprints at Laetoli, northern Tanzania. Nature, 290, 496–498. Charteris, J., Wall, J. C., & Nottrodt, J. W. (1982). Pliocene hominid gait: New interpretations based on available footprint data from Laetoli. American Journal of Physical Anthropology, 58, 133–144. Clarke, R. J. (1979). Early hominid footprints from Tanzania. South African Journal of Science, 75, 148–149. Clarke, R. J. (1988). A new Australopithecus cranium from Sterkfontein and its bearing on the ancestry of Paranthropus. In F. E. Grine (Ed.), Evolutionary history of the “Robust” Australopithecines (pp. 285– 292). New York: Aldine de Gruyter. Cole, T. M., & Smith, F. H. (1987). An odontometric assessment of variability in Australopithecus afarensis. Human Evolution, 2, 221–234. Constantino, P., & Wood, B. (2007). The evolution of Zinjanthropus boisei. Evolutionary Anthropology, 16, 49–62. Dart, R. A. (1925). Australopithecus africanus: The man-ape of South Africa. Nature, 115, 195–199. Day, M. H., & Wickens, E. H. (1980). Laetoli Pliocene hominid footprints and bipedalism. Nature, 286, 385–387. Day, M. H., Leakey, R. E. F., Walker, A. C., & Wood, B. A. (1976). New hominids from East Turkana, Kenya. American Journal of Physical Anthropology, 45, 369–436. Day, M. H., Leakey, M. D., & Olson, T. R. (1980a). On the status of Australopithecus afarensis. Science, 207, 1102–1103. Day, M. H., Leakey, M. D., & Magori, C. (1980b). A new hominid fossil skull (L.H. 18) from the Ngaloba Beds, Laetoli, northern Tanzania. Nature, 284, 55–56. Deino, A. (2011). 40Ar/39Ar dating of Laetoli, Tanzania. In T. Harrison (Ed.), Paleontology and geology of Laetoli: Human evolution in context (Geology, geochronology, paleoecology and paleoenvironment, Vol. 1, pp. 77–97). Dordrecht: Springer.
T. Harrison Ditchfield, P., & Harrison, T. (2011). Sedimentology, lithostratigraphy and depositional history of the Laetoli area. In T. Harrison (Ed.), Paleontology and geology of Laetoli: Human evolution in context (Geology, geochronology, paleoecology and paleoenvironment, Vol. 1, pp. 47–76). Dordrecht: Springer. Drake, R., & Curtis, G. H. (1987). K-Ar geochronology of the Laetoli fossil localities. In M. D. Leakey & J. M. Harris (Eds.), Laetoli: A Pliocene site in northern Tanzania (pp. 48–52). Oxford: Oxford University Press. Dugan, J., & Holliday, T. W. (2009). Utility of the lateral meniscus notch in distinguishing hominin taxa. Journal of Human Evolution, 57, 773–776. Duncan, A. S., Kappelman, J., & Shapiro, L. J. (1994). Metatarso phalangeal joint function and positional behavior in Australopithecus afarensis. American Journal of Physical Anthropology, 93, 67–82. Falk, D., Gage, T. B., Dudek, B., & Olson, T. R. (1995). Did more than one species of hominid coexist before 3.0 Ma? Evidence from blood and teeth. Journal of Human Evolution, 29, 591–600. Feibel, C. S., Brown, F. H., & McDougall, I. (1989). Stratigraphic context of fossil homininds from the Omo Group deposits: Northern Turkana Basin, Kenya and Ethiopia. American Journal of Physical Anthropology, 78, 595–622. Feibel, C. S., Agnew, N., Latimer, B., Demas, M., Marshall, F., Waane, S. A. C., & Schmid, P. (1996). The Laetoli hominid footprints – a preliminary report on the conservation and scientific restudy. Evolutionary Anthropology, 4, 149–154. Ferguson, W. W. (1983). An alternative interpretation of Australopithecus afarensis fossil material. Primates, 24, 397–409. Ferguson, W. W. (1984). Revision of fossil hominid jaws from the PlioPleistocene of Hadar, in Ethiopia including a new species of the genus Homo (Hominoidea: Homininae). Primates, 25, 519–529. Ferguson, W. W. (1986). The taxonomic status of Praeanthropus africanus (Primates: Pongidae) from the Late Pliocene of eastern Africa. Primates, 27, 485–492. Ferguson, W. W. (1987). Revision of the subspecies of Australopithecus africanus (Primates: Hominidae), including a new subspecies from the late Pliocene of Ethiopia. Primates, 28, 258–265. Ferguson, W. W. (1988). Model of a hypothetical protohominid dentition and its implications for hominid phylogeny. Primates, 29, 535–547. Ferguson, W. W. (1989). A new species of the genus Australopithecus (Primates: Hominidae) from Plio/Pleistocene deposits West of Lake Turkana in Kenya. Primates, 30, 223–232. Grausz, H. M., Leakey, R. E., Walker, A. C., & Ward, C. V. (1988). Associated cranial and postcranial bones of Australopithecus boisei. In F. E. Grine (Ed.), Evolutionary history of the “Robust” Australopithecines (pp. 127–132). New York: Gruyter. Gray, J. E. (1825). An outline of an attempt at the disposition of Mammalia into tribes and families, with a list of the genera apparently appertaining to each tribe. Annals of Philosophy, 10, 337–344. Gregory, W. K. (Ed.). (1950). The anatomy of the gorilla. New York: Columbia University Press. Grine, F. E., Ungar, P. S., Teaford, M. F., & El-Zaatari, S. (2006). Molar microwear in Praeanthropus afarensis: Evidence for dietary stasis through time and under diverse paleoecological conditions. Journal of Human Evolution, 51, 297–319. Groves, C. P. (1996). Case 2998. Australopithecus afarensis Johanson, 1978 (Mammalia, Primates): Proposed conservation of the specific name. Bulletin of Zoological Nomenclature, 53, 24–27. Haile-Selassie, Y., Deino, A., Saylor, B., Umer, M., & Latimer, B. (2007). Preliminary geology and paleontology of new hominidbearing Pliocene site in the central Afar region of Ethiopia. Anthropological Science, 115, 215–222. Haile-Selassie, Y., Saylor, B. Z., Deino, A., Alene, M., & Latimer, B. M. (2010). New hominid fossils from Woranso-Mille (Central Afar,
7 Hominins from Laetoli Ethiopia) and taxonomy of early Australopithecus. American Journal of Physical Anthropology, 141, 406–417. Harris, J. W. K., & Harris, K. (1981). A note on the archaeology at Laetoli. Nyame Akuma, 189, 18–21. Harrison, T. (1993). Cladistic concepts and the species problem in hominoid evolution. In W. H. Kimbel & L. B. Martin (Eds.), Species, species concepts, and primate evolution (pp. 345–371). New York: Plenum. Harrison, T. (2002). The first record of fossil hominins from the Ndolanya Beds, Laetoli, Tanzania. American Journal of Physical Anthropology Supplement, 32, 83. Harrison, T., & Kweka, A. (2011). Paleontological localities on the Eyasi Plateau, including Laetoli. In T. Harrison (Ed.), Paleontology and geology of Laetoli: Human evolution in context (Geology, geochronology, paleoecology and paleoenvironment, Vol. 1, pp. 17–45). Dordrecht: Springer. Hashemi, J., Chandrashekar, N., Gill, B., Beynnon, B. D., Slauterbeck, J. R., Schutt, R. C., Mansouri, H., & Dabezies, E. (2008). The geometry of the tibial plateau and its influence on the biomechanics of the tibiofemoral joint. The Journal of Bone and Joint Surgery, 90, 2724–2734. Hay, R. L. (1987). Geology of the Laetoli area. In M. D. Leakey & J. M. Harris (Eds.), Laetoli: A Pliocene site in northern Tanzania (pp. 23–47). Oxford: Clarendon. Hay, R. L., & Leakey, M. D. (1982). The fossil footprints of Laetoli. Scientific American, 246, 50–57. Hennig, E. (1948). Quartärfaunen und Urgeschichte Ostafrikas. Naturwissenschaftliche Rundschau, 1, 212–217. Hill, A. (1987). Damage to some fossil bones from Laetoli. In M. D. Leakey & J. M. Harris (Eds.), Laetoli: A Pliocene site in northern Tanzania (pp. 543–545). Oxford: Clarendon. Hill, A. (1999). The Baringo Basin, Kenya: From Bill Bishop to BPRP. In P. Andrews & P. Banham (Eds.), Late Cenozoic environments and hominid evolution: A tribute to Bill Bishop (pp. 85–97). London: Geological Society of London. Hinrichsen, D. (1978). How old are our ancestors. New Scientist, 78, 571. Hlusko, L. J. (2004). Protostylid variation in Australopithecus. Journal of Human Evolution, 46, 579–594. Howell, F. C. (1978). Hominidae. In V. J. Maglio & H. B. S. Cooke (Eds.), Evolution of African mammals (pp. 154–248). Cambridge: Harvard University Press. Johanson, D. C. (1980). Early African hominid phylogenesis: A re-evaluation. In L.-K. Königsson (Ed.), Current argument on early man (pp. 31–69). Oxford: Pergamon. Johanson, D. C., & Coppens, Y. (1976). A preliminary anatomical diagnosis of the first Plio/Pleistocene hominid discoveries in the Central Afar, Ethiopia. American Journal of Physical Anthropology, 45, 217–234. Johanson, D. C., & Taieb, M. (1976). Plio-Pleistocene hominid discoveries in Hadar, Ethiopia. Nature, 260, 293–297. Johanson, D. C., White, T. D., & Coppens, Y. (1978). A new species of the genus Australopithecus (Primates: Hominidae) from the Pliocene of eastern Africa. Kirtlandia, 28, 1–14. Johanson, D. C., White, T. D., & Coppens, Y. (1982). Dental remains from the Hadar Formation, Ethiopia: 1974–1977 collections. American Journal of Physical Anthropology, 57, 545–603. Jungers, W. L. (1982). Lucy’s limbs: Skeletal allometry and locomotion in Australopithecus afarensis. Nature, 297, 676–678. Jungers, W. L. (1988). New estimates of body size in australopithecines. In F. E. Grine (Ed.), Evolutionary history of the “robust” australopithecines (pp. 115–125). New York: Aldine de Gruyter. Jungers, W. L. (1991). A pygmy perspective on body size and shape in Australopithecus afarensis (A.L. 288-1, “Lucy”). In B. Senut & Y. Coppens (Eds.), Origine(s) de la bipédie chez les Hominidés (pp. 215–224). Paris: CNRS. Jungers, W. L., & Stern, J. T. (1983). Body proportions, skeletal allometry and locomotion in the Hadar hominids: A reply to Wolpoff. Journal of Human Evolution, 12, 673–684.
185 Kaiser, T., Bromage, T. G., & Schrenk, F. (1995). Hominid Corridor Research Project update: New Pliocene fossil localities at Lake Manyara and putative oldest Early Stone Age occurrences at Laetoli (Upper Ndolanya Beds), northern Tanzania. Journal of Human Evolution, 28, 117–120. Keyser, A. W. (2000). The Drimolen skull: The most complete australopithecine cranium and mandible to date. South African Journal of Science, 96, 189–193. Kibunjia, M., Roche, H., Brown, F. H., & Leakey, R. E. F. (1992). Pliocene and Pleistocene archaeological sites west of Lake Turkana, Kenya. Journal of Human Evolution, 23, 431–438. Kimbel, W. H. (2009). The origins of Homo. In F. E. Grine, J. G. Fleagle, & R. E. Leakey (Eds.), The first humans: Origin and early evolution of the genus Homo (pp. 31–37). Dordrecht: Springer. Kimbel, W. H., & Delezene, L. K. (2009). “Lucy” redux: A review of research on Australopithecus afarensis. Yearbook of Physical Anthropology, 52, 2–48. Kimbel, W. H., White, T. D., & Johanson, D. C. (1985). Craniodental morphology of the hominids from Hadar and Laetoli: Evidence of “Paranthropus” and Homo in the mid-Pliocene of eastern Africa. In E. Delson (Ed.), Ancestors: The hard evidence (pp. 120–137). New York: Alan R. Liss. Kimbel, W. H., White, T. D., & Johanson, D. C. (1988). Implications of KNM-WT 17000 for the evolution of the “robust” australopithecus. In F. E. Grine (Ed.), Evolutionary history of the “Robust” Australopithecines (pp. 259–268). New York: Gruyter. Kimbel, W. H., Walter, R. C., Johanson, D. C., Reed, K. E., Aronson, J. L., Assefa, Z., Marean, C. W., Eck, G. G., Bobe, R., Hovers, E., Rak, Y., Vondra, C., Yemane, T., York, D., Chen, Y., Evensen, N. M., & Smith, P. E. (1996). Late Pliocene Homo and Oldowan tools from the Hadar Formation (Kada Hadar Member), Ethiopia. Journal of Human Evolution, 31, 549–561. Kimbel, W. H., Rak, Y., & Johanson, D. C. (2004). The skull of Australopithecus afarensis. Oxford: Oxford University Press. Kimbel, W. H., Lockwood, C. A., Ward, C. V., Leakey, M. G., Rak, Y., & Johanson, D. C. (2006). Was Australopithecus anamensis ancestral to A. afarensis? A case of anagenesis in the hominin fossil record. Journal of Human Evolution, 51, 134–152. Kohl-Larsen, L. (1943). Auf den Spuren des Vormenschen (2nd Band). Stuttgart: Strecker und Schröder. Kullmer, O., Sandrock, O., Abel, R., Schrenck, F., Bromage, T. G., & Juwayeyi, Y. M. (1999). The first Paranthropus from the Malawi Rift. Journal of Human Evolution, 37, 121–127. Kyauka, P. S. (1994). A comparative study of the Laetoli Hominid 21 skeleton and its implications for the developmental biology of Australopithecus afarensis. Ph.D. dissertation, University of California, Berkeley. Kyauka, P. S., & Ndessokia, P. (1990). A new hominid tooth from Laetoli, Tanzania. Journal of Human Evolution, 19, 747–750. Latimer, B. (1991). Locomotor adaptations in Australopithecus afarensis: The issue of arboreality. In B. Senut & Y. Coppens (Eds.), Origine(s) de la Bipédie chez les Hominidés (pp. 169–176). Paris: CNRS. Latimer, B., Ohman, J. C., & Lovejoy, C. O. (1987). Talocrual joint in African hominoids: Implications for Australopithecus afarensis. American Journal of Physical Anthropology, 74, 155–175. Leakey, M. D. (1978). Pliocene footprints at Laetoli, northern Tanzania. Antiquity, 52, 133. Leakey, M. D. (1979). 3.6 million years old footprints in the ashes of time. National Geographic, 155, 446–457. Leakey, M. D. (1981). Tracks and tools. Philosophical Transactions of the Royal Society of London, Series B, 292, 95–102. Leakey, M. D. (1987a). Introduction. In M. D. Leakey & J. M. Harris (Eds.), Laetoli: A Pliocene site in northern Tanzania (pp. 1–22). Oxford: Clarendon.
186 Leakey, M. D. (1987b). The Laetoli hominid remains. In M. D. Leakey & J. M. Harris (Eds.), Laetoli: A Pliocene site in northern Tanzania (pp. 108–117). Oxford: Clarendon. Leakey, M. D. (1987c). The hominid footprints: Introduction. In M. D. Leakey & J. M. Harris (Eds.), Laetoli: A Pliocene site in northern Tanzania (pp. 490–496). Oxford: Clarendon. Leakey, M. D., & Hay, R. L. (1979). Pliocene footprints in the Laetolil Beds at Laetoli, northern Tanzania. Nature, 278, 317–323. Leakey, R. E. F., & Walker, A. (1988). New Australopithecus boisei specimens from East and West Lake Turkana, Kenya. American Journal of Physical Anthropology, 76, 1–24. Leakey, M. D., Hay, R. L., Curtis, G. H., Drake, R. E., Jackes, M. K., & White, T. D. (1976). Fossil hominids from the Laetolil Beds. Nature, 262, 460–466. Leakey, M. G., Feibel, C. S., McDougall, I., & Walker, A. (1995). New four-million-year-old hominid species from Kanapoi and Allia Bay, Kenya. Nature, 376, 565–571. Leakey, M. G., Feibel, C. S., McDougall, I., Ward, C., & Walker, A. (1998). New specimens and confirmation of an early age for Australopithecus anamensis. Nature, 393, 62–66. Le Minor, J. M. (1990). Comparative morphology of the lateral meniscus of the knee in primates. Journal of Anatomy, 170, 161–171. Lockwood, C. A., Kimbel, W. H., & Johanson, D. C. (2000). Temporal trends and metric variation in the mandibles and dentition of Australopithecus afarensis. Journal of Human Evolution, 39, 23–55. Logan, T. R., Lucas, S. G., & Sobus, J. C. (1983). The taxonomic status of Australopithecus afarensis Johanson in Hinrichsen 1978 (Mammalia, Primates). Haliksa’a: UNM Contributions to Anthropology, 2, 6–27. Lovejoy, C. O., Johanson, D. C., & Coppens, Y. (1982). Hominid lower limb bones recovered from the Hadar Formation: 1974–1977 collections. American Journal of Physical Anthropology, 57, 679–700. Lydekker, R. (1879). Further notices of Siwalik Mammalia. Record of the Geological Survey of India, 12, 33–52. Mabulla, A. Z. P. (2000). Strategy for Cultural Heritage Management (CHM) in Africa: A case study. African Archaeological Review, 17, 211–233. Magori, C. C., & Day, M. H. (1983). Laetoli Hominid 18: An early Homo sapiens skull. Journal of Human Evolution, 12, 747–753. Manega, P. (1993). Geochronology, geochemistry and isotopic study of the Plio-Pleistocene hominid sites and the Ngorongora volcanic highlands in northern Tanzania. Ph.D. dissertation, University of Colorado at Boulder, Boulder. McCollum, M. A. (1997). Palatal thickening and facial form in Paranthropus: Examination of alternative developmental models. American Journal of Physical Anthropology, 103, 375–392. McCollum, M. (1999). The robust australopithecine face: A morphometric perspective. Science, 284, 301–305. McCollum, M. A., Grine, F. E., Ward, S. C., & Kimbel, W. H. (1993). Subnasal morphological variation in extant hominoids and fossil hominids. Journal of Human Evolution, 24, 87–111. McHenry, H. M. (1986). The first bipeds: A comparison of the Australopithecus afarensis and Australopithecus africanus postcranium and implications for the evolution of bipedalism. Journal of Human Evolution, 15, 177–191. McHenry, H. M. (1991). First steps? Analyses of the postcranium of early hominids. In B. Senut & Y. Coppens (Eds.), Origine(s) de la Bipédie chez les Hominidés (pp. 133–142). Paris: CNRS. McHenry, H. M. (1994). Early hominid postcrania: Phylogeny and function. In R. S. Corruccini & R. L. Ciochon (Eds.), Integrative paths to the past: Paleoanthropological advances in honor of F. Clark Howell (pp. 251–268). Englewood Cliffs: Prentice Hall. Meldrum, D. J. (2004). Fossilized Hawaiian footprints compared with Laetoli hominid footprints. In D. J. Meldrum & C. E. Hilton (Eds.), From biped to strider: The emergence of modern human walking,
T. Harrison running and resource transport (pp. 63–83). New York: Kluwer/ Plenum. Mollel, G. F., Swisher, C. C., III, Feigenson, M. D., & Carr, J. D. (2011). Petrology, geochemistry and age of Satiman, Lemagurut and Oldeani: Sources of the volcanic deposits of the Laetoli area. In T. Harrison (Ed.), Paleontology and geology of Laetoli: Human evolution in context (Geology, geochronology, paleoecology and paleoenvironment, Vol. 1, pp. 99–119). Dordrecht: Springer. Ndessokia, P. N. S. (1990). The mammalian fauna and archaeology of the Ndolanya and Olpiro Beds, Laetoli, Tanzania. Ph.D. dissertation, University of California, Berkeley. Olson, T. R. (1981). Basicranial morphology of the extant hominoids and Pliocene hominids: The new material from the Hadar Formation, Ethiopia and its significance in early human evolution and taxonomy. In C. B. Stringer (Ed.), Aspects of human evolution (pp. 99–128). London: Taylor & Francis. Olson, T. R. (1985). Cranial morphology and systematics of the Hadar Formation hominids and “Australopithecus” africanus. In E. Delson (Ed.), Ancestors: The hard evidence (pp. 102–119). New York: Alan R. Liss. Organ, J. M., & Ward, C. (2006). Contours of the hominoid lateral tibial condyle with implications for Australopithecus. Journal of Human Evolution, 51, 113–127. Pickford, M. L. (2004). Southern Africa: A cradle of evolution. South African Journal of Science, 100, 205–214. Prat, S., Brugal, J.-P., Tiercelin, J.-J., Barrat, J.-A., Bohn, M., Delagnes, A., Harmand, S., Kimeu, K., Kibunjia, M., Texier, P.-J., & Roche, H. (2005). First occurrence of early Homo in the Nachukui Formation (West Turkana, Kenya) at 2.3–2.45 Myrs. Journal of Human Evolution, 49, 230–240. Protsch, R. (1976). The position of the Eyasi and Garusi hominids in East Africa. In P. V. Tobias & Y. Coppens (Eds.), Le Plus Anciens Hominidés, Colloque VI, Union Internationale des Sciences Préhistoriques et protohistoriqiues, IXe Congrés (pp. 207–237). Paris: CNRS. Protsch, R. R. R. (1981). The palaeoanthropological finds of the Pliocene and Pleistocene. In H. Müller-Beck (Ed.), Die Archäologischen und Anthropologischen Ergebnisse der KohlLarsen-Expeditionen in Nord-Tanzania 1933–1939 (Band 4, 3, pp. 1–181). Tübingen: Verlag Archaeologica Venatoria, Institute für Urgeschichte der Universität Tübingen. Puech, P.-F., Cianfarani, F., & Roth, H. (1986). Reconstruction of the maxillary dental arcade of Garusi Hominid 1. Journal of Human Evolution, 15, 325–332. Raichlen, D., Pontzer, H., & Sockol, M. D. (2008). The Laetoli footprints and early hominin locomotor kinematics. Journal of Human Evolution, 54, 112–117. Rak, Y. (1983). The australopithecine face. New York: Academic. Rak, Y. (1985). Systematic and functional implications of the facial morphology of Australopithecus and early Homo. In E. Delson (Ed.), Ancestors: The hard evidence (pp. 168–170). New York: Alan R. Liss. Rak, Y., Ginzburg, A., & Geffen, E. (2007). Gorilla-like anatomy on Australopithecus afarensis mandibles suggests Au. afarensis link to robust australopiths. Proceedings of the National Academy of Sciences USA, 104, 6568–6572. Remane, A. (1950). Die Zähne des Meganthropus africanus. Zeitschrift für Morphologie und Anthropologie, 42, 311–329. Remane, A. (1954). Structure and relationships of Meganthropus africanus. American Journal of Physical Anthropology, 12, 123–126. Robinson, J. T. (1953). Meganthropus, australopithecines and hominids. American Journal of Physical Anthropology, 11, 1–38. Robinson, J. T. (1954). The genera and species of the Australopithecinae. American Journal of Physical Anthropology, 12, 181–200. Robinson, J. T. (1955). Further remarks on the relationship between “Meganthropus” and australopithecines. American Journal of Physical Anthropology, 13, 429–445.
7 Hominins from Laetoli Robinson, C. A. (2003). Extant hominoid and australopith mandibular morphology; assessing alpha taxonomy and phylogeny in hominoids using mandibular characters. Ph.D. dissertation, New York University, New York. Roche, H., Blumenschine, R. J., & Shea, J. J. (2009). Origins and adaptations of early Homo: What archeology tells us. In F. E. Grine, J. G. Fleagle, & R. E. Leakey (Eds.), The first humans: Origin and early evolution of the genus Homo (pp. 135–147). Dordrecht: Springer. Sellers, W. I., Cain, G. M., Wang, W., & Crompton, R. H. (2005). Stride lengths, speed and energy costs in walking of Australopithecuds afarensis using evolutionary robotics to predict locomotion of early human ancestors. Journal of the Royal Society Interface, 2, 431–441. Semaw, S., Renne, R., Harris, J. W. K., Feibel, C. S., Bernor, R. L., Fessaha, N., & Mowbray, K. (1997). 2.5-million-year-old stone tools from Gona, Ethiopia. Nature, 385, 333–335. Semaw, S., Rogers, M. J., Quade, J., Renne, P. R., Butler, R. F., Dominguez-Rodrigo, M., Stout, D., Hart, W. S., Pickering, T., & Simpson, S. W. (2003). 2.6-million-year-old stone tools and associated bones from OGS-6 and OGS-7, Gona, Afar, Ethiopia. Journal of Human Evolution, 45, 169–177. Senut, B. (1980). New data on the humerus and its joints in PlioPleistocene hominids. Collegium Antropologicum, 4, 87–93. Senut, B. (1996). Pliocene hominid systematics and phylogeny. South African Journal of Science, 92, 165–166. Senut, B., & Tardieu, C. (1985). Functional aspects of Plio-Pleistocene hominid limb bones: Implications for taxonomy and phylogeny. In E. Delson (Ed.), Ancestors: The hard evidence (pp. 193–201). New York: Liss. Senyürek, M. (1955). A note on the teeth of Meganthropus africanus Weinert from Tanganyika Territory. Belleten, 19, 1–55. Skelton, R. R., & McHenry, H. M. (1992). Evolutionary relationships among early hominids. Journal of Human Evolution, 23, 309–349. Skelton, R. R., McHenry, H. M., & Drawhorn, G. M. (1986). Phylogenetic analysis of early hominids. Current Anthropology, 27, 21–43. Steininger, C., Berger, L. R., & Kuhn, B. F. (2008). A partial skull of Paranthropus robustus from Cooper’s Cave, South Africa. South African Journal of Science, 104, 143–146. Stern, J. T. (2000). Climbing to the top: A personal memoir of Australopithecus afarensis. Evolutionary Anthropology, 9, 113–133. Stern, J. T., & Susman, R. L. (1983). The locomotor anatomy of Australopithecus afarensis. American Journal of Physical Anthropology, 60, 279–317. Stern, J. T., & Susman, R. L. (1991). “Total morphological pattern” versus the “magic trait”: Conflicting approaches to the study of early hominid bipedalism. In B. Senut & Y. Coppens (Eds.), Origine(s) de la bipédie chez les Hominidés (pp. 99–111). Paris: CNRS. Strait, D. S., & Grine, F. E. (2001). The systematics of Australopithecus garhi. Ludus Vitalis, 9, 109–135. Strait, D. S., & Grine, F. E. (2004). Inferring hominoid and early hominid phylogeny using craniodental characters: The role of fossil taxa. Journal of Human Evolution, 47, 399–452. Strait, D. S., Grine, F. E., & Moniz, M. A. (1997). A reappraisal of early hominid phylogeny. Journal of Human Evolution, 32, 17–82. Su, D., & Harrison, T. (2008). Ecological implications of the relative rarity of fossil hominins at Laetoli. Journal of Human Evolution, 55, 672–681. Susman, R. L., & Stern, J. T. (1991). Locomotor behavior of early hominids: Epistemology and fossil evidence. In Y. Coppens & B. Senut (Eds.), Origine(s) de la Bipédie chez les Hominidés, Cahiers de Paléoanthropologie (pp. 121–132). Paris: CNRS. Susman, R. L., Stern, J. T., & Jungers, W. L. (1984). Arboreality and bipedality in the Hadar hominids. Folia Primatologica, 43, 113–156.
187 Susman, R. L., Stern, J. T., & Jungers, W. L. (1985). Locomotor adaptations in the Hadar hominids. In E. Delson (Ed.), Ancestors: The hard evidence (pp. 184–192). New York: Alan R. Liss. Suwa, G. (1988). Evolution of the “robust” australopithecines in the Omo Succession: Evidence from mandibular premolar morphology. In F. E. Grine (Ed.), Evolutionary history of the “robust” australopithecines (pp. 199–222). New York: Gruyter. Suwa, G. (1989). The premolar of KNM-WT 17000 and relative anterior to posterior dental size. Journal of Human Evolution, 18, 795–799. Suwa, G., White, T. D., & Howell, F. C. (1996). Mandibular postcanine dentition from the Shungura Formation, Ethiopia: Crown morphology, taxonomic allocations, and Plio-Pleistocene hominid evolution. American Journal of Physical Anthropology, 101, 247–282. Suwa, G., Asfaw, B., Beyene, Y., White, T. D., Katoh, S., Nagaoka, S., Nakaya, H., Uzawa, K., Renne, P., & WoldeGabriel, G. (1997). The first skull of Australopithecus boisei. Nature, 389, 489–492. Tardieu, C. (1982). Caractères plésiomorphes et apomorphes de l’articulation du genou chez les primates hominoids. Geobios, mémoire special, 6, 321–334. Tardieu, C. (1983). L’articulation du genou des primates catarhiniens et hominids fossils. Implications phylogenetique et taxinomique. Bulletin et mémoires de la Société d’Anthropologie de Paris, 10(Série XIII), 355–372. Tardieu, C. (1986a). The knee joint in three hominoid primates: Application to Plio-Pleistocene hominids and evolutionary implications. In D. Taub & F. King (Eds.), Current perspectives in primate biology (pp. 182–190). New York: Van Nostrand. Tardieu, C. (1986b). Evolution of the knee intra-articular menisci in primates and some fossil hominids. In J. G. Else & P. C. Lee (Eds.), Primate evolution. Proceedings of the 10th congress of the international primatological society (Vol. 1, pp. 183–190). Cambridge: Cambridge University Press. Tardieu, C. (1999). Ontogeny and phylogeny of femoro-tibial characters in humans and hominid fossils: Functional influence and genetic determinism. American Journal of Physical Anthropology, 110, 365–377. Tobias, P. V. (1967). The cranium and maxillary dentition of Australopithecus (Zinjanthropus) boisei. Olduvai Gorge (Vol. 2). Cambridge: Cambridge University Press. Tobias, P. V. (1980a). A survey and synthesis of the African hominids of the late Tertiary and early Quaternary periods. In L.-K. Königsson (Ed.), Current argument on early man (pp. 86–113). Oxford: Pergamon. Tobias, P. V. (1980b). “Australopithecus afarensis” and A. africanus: Critique and an alternative hypothesis. Palaeontologia Africana, 23, 1–17. Tobias, P. V. (1991). Olduvai Gorge (The skulls, endocasts and teeth of Homo habilis, Vol. 4). Cambridge: Cambridge University Press. Tuttle, R. H. (1985). Ape footprints and Laetoli impressions: A response to the SUNY claims. In P. V. Tobias (Ed.), Hominid evolution: Past, present and future (pp. 129–133). New York: Alan R. Liss. Tuttle, R. H. (1987). Kinesiological inference and evolutionary implications from Laetoli bipedal trails G-1, G-2/3 and A. In M. D. Leakey & J. M. Harris (Eds.), Laetoli: A Pliocene site in northern Tanzania (pp. 503–523). Oxford: Clarendon. Tuttle, R. H. (1990). The pitted pattern of Laetoli feet. Natural History, 90, 60–65. Tuttle, R. H. (1994). Up from electromyography: Primate energetics and the evolution of human bipedalism. In R. S. Corruccini & R. L. Ciochon (Eds.), Integrative paths to the past (pp. 251–268). Englewood Cliffs: Prentice Hall. Tuttle, R. H. (2008). Footprint clues in hominid evolution and forensics: Lessons and limitations. Ichnos, 15, 158–165.
188 Tuttle, R. H., Webb, D. M., Weidl, E., & Baksh, M. (1990). Further progress on the Laetoli trails. Journal of Archaeological Science, 17, 347–362. Tuttle, R. H., Webb, D. M., & Baksh, M. (1991a). Laetoli toes and Australopithecus afarensis. Human Evolution, 6, 193–200. Tuttle, R. H., Webb, D., & Tuttle, N. (1991b). Laetoli footprint trails and the evolution of hominid bipedalism. In Y. Coppens & B. Senut (Eds.), Origine(s) de la Bipédie chez les Hominidés. Cahiers de Paléoanthropologie (pp. 203–218). Paris: CNRS. Tuttle, R. H., Webb, D. M., Tuttle, N. I., & Baksh, M. (1992). Footprints and gaits of bipedal apes, bears, and barefoot people: Perspective on Pliocene tracks. In S. Matano, R. H. Tuttle, H. Ishida, & M. Goodman (Eds.), Topics in primatology (Evolutionary biology, reproductive endocrinology, and virology, Vol. 3, pp. 221–242). Tokyo: University of Tokyo Press. Ullrich, H. (2001). Garusi Hominid 4 – Ein Australoppithecinenzahn aus der Sammlung Kohl-Larsen. In M. Schultz, K. Atzwanger, G. Bräuer, K. Christiansen, J. Forster, H. Greil, W. Henke, U. Jaeger, C. Niemitz, C. Scheffler, W. Schiefenhövel, I. Schröder, & I. Weichmann (Eds.), Homo – Unsere Herkunft und Kukunft: Proceedings – 4. Kongress der Gesellschaft für Anthropologie e.V (pp. 52–54). Göttingen: Cuvillier Verlag. Walker, A., & Leakey, R. E. (1988). The evolution of Australopithecus boisei. In F. E. Grine (Ed.), Evolutionary history of the “Robust” Australopithecines (pp. 247–258). New York: Gruyter. Walker, A., Leakey, R. E., Harris, J. M., & Brown, F. H. (1986). 2.5Myr Australopithecus boisei from west of Lake Turkana, Kenya. Nature, 322, 517–522. Ward, C. V. (2002). Interpreting the posture and locomotion of Australopithecus afarensis: Where do we stand? Yearbook of Physical Anthropology, 45, 185–215. Ward, S., Johanson, D. C., & Coppens, Y. (1982). Subocclusal morphology and alveolar process relationships of hominid gnathic elements from the Hadar Formation: 1974–1977 collections. American Journal of Physical Anthropology, 57, 605–630. Ward, C., Leakey, M. G., & Walker, A. (1999). The new hominid species Australopithecus anamensis. Evolutionary Anthropology, 7, 197–205. Ward, C. V., Leakey, M. G., & Walker, A. (2001). Morphology of Australopithecus anamensis from Kanapoi and Allia Bay, Kenya. Journal of Human Evolution, 41, 255–368. Ward, S. C., & Kimbel, W. H. (1983). Subnasal morphology and the systematic position of Sivapithecus. American Journal of Physical Anthropology, 61, 157–171.
T. Harrison Weinert, H. (1950). Über die neuen Vor- und Frühmenschenfunde aus Afrika, Java, China und Frankreich. Zeitschrift für Morphologie und Anthropologie, 42, 113–145. White, T. D. (1977). New fossil hominids from Laetolil, Tanzania. American Journal of Physical Anthropology, 46, 197–230. White, T. D. (1980a). Additional fossil hominids from Laetoli, Tanzania: 1976–1979 specimens. American Journal of Physical Anthropology, 53, 487–504. White, T. D. (1980b). Evolutionary implications of Pliocene hominid footprints. Science, 208, 175–176. White, T. D. (1981). Primitive hominid canine from Tanzania. Science, 213, 348–349. White, T. D. (1985). The hominids of Hadar and Laetoli: An elementby-element comparison of the dental samples. In E. Delson (Ed.), Ancestors: The hard evidence (pp. 138–152). New York: Liss. White, T. D., & Suwa, G. (1987). Hominid footprints at Laetoli: Facts and interpretations. American Journal of Physical Anthropology, 72, 485–514. White, T. D., Johanson, D. C., & Kimbel, W. H. (1981). Australopithecus africanus: Its phyletic position reconsidered. South African Journal of Science, 77, 445–470. White, T. D., Suwa, G., Simpson, S., & Asfaw, B. (2000). Jaws and teeth of Australopithecus afarensis from Maka, Middle Awash, Ethiopia. American Journal of Physical Anthropology, 111, 45–68. White, T. D., WoldeGabriel, G., Asfaw, B., Ambrose, S., Beyene, Y., Bernor, R. L., Boisserie, J. R., Currie, B., Gilbert, H., Haile-Selassie, Y., Hart, W. K., Hlusko, L. J., Howell, F. C., Kono, R. T., Lehmann, T., Louchart, A., Lovejoy, C. O., Renne, P. R., Saegusa, H., Vrba, E., Wesselman, H., & Suwa, G. (2006). Asa Issie, Aramis and the origin of Australopithecus. Nature, 440, 883–889. Wolpoff, M. H. (1999). Paleoanthropology (2nd ed.). Boston: McGraw-Hill. Wood, B. A. (1988). Are ‘robust’ australopithecines a monophyletic group? In F. E. Grine (Ed.), Evolutionary history of the “robust” australopithecines (pp. 269–284). New York: Gruyter. Wood, B. A. (1991). Koobi Fora research project (Hominid cranial remains, Vol. 4). Oxford: Clarendon. Wood, B., & Constantino, P. (2007). Paranthropus boisei: Fifty years of evidence and analysis. Yearbook of Physical Anthropology, 50, 106–132. Wood, B. A., & Richmond, B. G. (2000). Human evolution: Taxonomy and paleobiology. Journal of Anatomy, 196, 19–60.
Chapter 8
Carnivora Lars Werdelin and Reihaneh Dehghani
Abstract This paper reviews the extensive carnivoran fauna of Laetoli on the basis of collections housed in Berlin, London, Nairobi, and Dar es Salaam. Members of the Carnivora are known from both the Lower and Upper Laetolil Beds, as well as from the Upper Ndolanya Beds. Of these, the Upper Laetolil Beds are best sampled, and the material includes a minimum of 28 species of Carnivora (four Canidae, three Mustelidae, three Viverridae, six Herpestidae, five Hyaenidae, and seven Felidae). Many of the smaller Carnivora species include complete or partial skeletons and whole, undamaged crania, suggesting rapid burial and absence of trampling and other taphonomic processes that severely affected the more fragmentary larger Carnivora. The Upper Ndolanya Beds Carnivora are preserved in a similar fashion. This stratigraphic unit includes nine to ten species (one Mustelidae, two Herpestidae, one or two Hyaenidae, and five Felidae). All of these are also known from the Upper Laetolil Beds. The Lower Laetolil Beds are less well sampled, with only four species of Carnivora (one Mustelidae, one Herpestidae, and two Hyaenidae). Of these, the mustelid and one hyenid are unique to this stratigraphic unit, while one hyenid is shared with the Upper Laetolil Beds and the herpestid with both the Upper Laetolil Beds and the Upper Ndolanya Beds. Three of the Lower Laetolil Beds Carnivora (all except the herpestid) are partial skeletons, suggesting different depositional or taphonomic conditions at that time, while the presence of an otter in the Lower Laetolil Beds indicates the presence of a large, permanent body of water in the vicinity. Keywords Canidae • Felidae • Herpestidae • Hyaenidae • Mustelidae • Viverridae • Pliocene • Laetoli
L. Werdelin (*) and R. Dehghani Department of Palaeozoology, Swedish Museum of Natural History, Box 50007 S-104 05, Stockholm, Sweden e-mail:
[email protected];
[email protected] Introduction Carnivorans from the Laetoli deposits have been described several times since their first discovery in the 1930s. The first treatment was by Dietrich (1942). The exact provenance of most of the material described by Dietrich is not known with certainty, but specimens recorded as coming from korongos (gullies) in the Garussi (now spelled “Garusi”) region are generally considered to come from the Upper Laetolil Beds. Dietrich described the following taxa from these beds: Canis (Lupulella) mesomelas ssp., Mungos palaeoserengetensis n. sp., Mungos palaeogracilis n. sp., Crocuta crocuta subsp., and Panthera pardus subsp. (a Felis of the “ocreata group” is also recorded from Garussi, but these specimens are explicitly stated to be subfossil). Of the Laetolil Beds carnivorans recovered by the German expedition under the direction of Kohl-Larsen (described in Kohl-Larsen [1943]), only a skull and left horizontal mandibular ramus of M. palaeoserengetensis, Garussi 2/39, were figured by Dietrich (1942, Plate IV, figs. 31 and 36). Some 20 years later, Petter (1963) published a study of the “viverrids” (= Viverridae and Herpestidae) of Laetoli. In this work, Petter reassigned M. palaeoserengetensis and M. palaeogracilis to Herpestes and described a new species of herpestid, Mungos dietrichi. In Petter’s conception, H. palaeoserengetensis is based on the skull illustrated by Dietrich (1942, Pl. IV, fig. 31), whereas M. dietrichi is based on the mandible that Dietrich associated with the skull and illustrated in his Pl. IV, fig. 36. In addition, Petter (1963) suggested the presence of a second species of Mungos on the basis of a skull fragment recovered in 1959 by Louis and Mary Leakey, but she later reassigned this specimen to M. dietrichi (Petter 1987). A second new species described by Petter (1963) is the very large viverrid Viverra leakeyi, one of a number of extinct viverrids that were considerably larger than any living member of the family. Collections at Laetoli by teams led by Mary Leakey in 1975–1981 (Leakey and Harris 1987) recovered large quantities of material, including nearly 600 specimens assigned to the Carnivora. These specimens were described by Petter (1987) (Herpestidae, Viverridae, Mustelidae, and two species
T. Harrison (ed.), Paleontology and Geology of Laetoli: Human Evolution in Context. Volume 2: Fossil Hominins and the Associated Fauna, Vertebrate Paleobiology and Paleoanthropology, DOI 10.1007/978-90-481-9962-4_8, © Springer Science+Business Media B.V. 2011
189
190
of Canidae) and Barry (1987) (Hyaenidae, Felidae, and remaining Canidae). The study by Petter (1987) constituted a revision of her previous work on the small carnivorans of Laetoli, which had been based almost entirely on the material collected by the Kohl-Larsen expeditions. Amongst the more important revisions in this work was reassigning Herpestes palaeogracilis to the genus Helogale. The study by Barry (1987) was the first detailed analysis of large carnivorans from Laetoli and one of the first studies of early Pliocene large carnivorans of eastern Africa. Although no new species-level taxa were described by Petter (1987) or Barry (1987), these studies did add numerous carnivoran taxa to the Laetoli fauna. Since the publications of Petter (1987) and Barry (1987), some minor revisions of the Laetoli carnivorans have taken place. Petter and Howell (1989) redescribed the Crocuta n. sp. of Barry (1987) and named it Crocuta dietrichi. Turner (1990) reassigned the Leo (=Panthera) aff. gombaszoegensis to Panthera leo, and Werdelin and Lewis (2001) reassigned the Megantereon sp. material from Laetoli to Dinofelis petteri. Recently, Hemmer et al. (2004) suggested that the Panthera cf. pardus from Laetoli shows affinities with Puma pardoides from the Pliocene of Eurasia and was interpreted by them as a primitive cougar rather than as a leopard. Most recently, Werdelin and Lewis (2005), in their review of eastern African Carnivora, implicitly revised the carnivoran faunal list from Laetoli. The justification for this revision is provided herein.
Material and Methods Laetoli carnivorans are housed in a number of institutions. The material collected by the Kohl-Larsen expeditions in the 1930s is housed in the Museum für Naturkunde, Berlin, Germany. Material collected in the 1950s by occasional expeditions and by the Mary Leakey expeditions of 1975– 1982 is housed in the Kenya National Museums, Nairobi, with a very small number housed in the Natural History Museum, London. Material collected by the Eyasi Plateau Research Project directed by Terry Harrison is housed in the National Museum of Tanzania, Dar es Salaam. We have studied all this material in preparation for writing this chapter. The carnivoran material from Laetoli is extensive and currently stands at 936 numbered specimens (excluding the material housed in Berlin and London). Most of these are fragmentary, rendering species-level identification difficult. A total of 496 of these specimens are cranial and/or dental, and these will form the focus of this contribution. Isolated, fragmentary postcranial bones are difficult to identify, and this will here be attempted only in special cases.
L. Werdelin and R. Dehghani
Some postcrania are associated with craniodental material, especially in the case of the smaller carnivorans, which in many cases are known from partial skeletons. Primary identifications have been based on visual inspection in the field and laboratory, complemented whenever possible with quantitative analysis. These analyses have been confined to bivariate methods, as the fragmentary nature of the material makes multivariate analysis impractical. In the following, material from pre-1975 Leakey expeditions has the prefix “LIT,” and material from the 1975–1981 expeditions is prefixed “LAET.” In both cases, the first two digits indicate the year of recovery (except as noted). Material from the Eyasi Plateau Research Project has the prefix “EP,” and the last two digits indicate the year of recovery. Other catalog number prefixes are “MB” (Museum für Naturkunde, Humboldt Universität, Berlin) and “NHM” (The Natural History Museum, London). In Tables 8.1–8.8, the following abbreviations are used: L, tooth length; W, tooth width; c, lower canine; p and P, lower and upper premolars, respectively; and m and M, lower and upper molars, respectively. Special measurements are as follows: Lpp4, length of main cusp of p4; Ltm1, length of trigonid of m1; LeP4, buccal length of P4; LiP4, lingual length of P4 (including protocone); WaP4, anterior width of P4; WblP4, width of anterior part of metastyle of P4; LpP4, length of paracone of P4; LmP4, length of metastyle of P4; C-C, width of snout between lateral sides of canines; P-P, width of palate between buccal sides of P4 metastyle; IOB, minimum interorbital width; POC, minimum postorbital constriction width; and ZB, maximum width of zygomatic arch.
Systematic Paleontology Order Carnivora Bowdich, 1821
Family Canidae Fischer, 1817 The Canidae are the least well known of the larger carnivorans in the fossil record of Africa. They are present in most faunas but are usually represented by a smaller number of specimens than, e.g., Hyaenidae or Felidae. It may be that the favored habitats of the majority of African Canidae are less well represented in the fossil record than those of other carnivoran families. Nevertheless, it is likely Africa holds a key position in the fossil record of Canidae, since the currently oldest published Canis is from this continent (Werdelin and Lewis 2000). Metric data for Laetoli Canidae are given in Tables 8.1 and 8.2.
LAET 75-3522 3.9 ?Nyctereutes barryi EP 1319/98 ?Nyctereutes barryi EP 227/05 ?Nyctereutes barryi EP 286/05 ?Nyctereutes barryi EP 597/01 ?Nyctereutes barryi EP 892/05 ?Nyctereutes barryi LAET 74-249 ?Nyctereutes barryi LAET 76-3844 ?Nyctereutes barryi EP 2431/03 cf. Canis sp. A EP 2126/00 aff. Otocyon sp. 2.4 For explanation of abbreviations see Material and Methods 5.0
2.0
Lp2 5.9
2.6
Table 8.1 Measurements of the lower dentition of Laetoli Canidae Catalog no. Taxon Lp1 Wp1 Wp2
2.8
2.9
3.2
6.6 2.7
3.6
7.6
5.6
2.8
6.5
Wp3 3.1
Lp3 7.1
11.6 6.1
5.3 3.2
4.0
7.7
Wp4 3.9
Lp4 8.5
Lm1
14.4
13.9
14.2 16.0 13.7
Wm1
6.2
6.5
6.4 6.4 6.5
9.4
9.4 11.4 9.6
Ltm1
Lm2 8.1
Wm2 6.0
8 Carnivora 191
?Nyctereutes barryi cf. Canis sp. A cf. Canis sp. A cf. Canis sp. A cf. Canis sp. A aff. Otocyon sp. aff. Otocyon sp. 4.9
3.3
For explanation of abbreviations see Material and Methods
LAET 75-3522 EP 1047/98 EP 437a/98 EP 437c/98 EP 437b/98 EP 1245/01 EP 208/01 10.9
5.7
Table 8.2 Measurements of the upper dentition of Laetoli Canidae Catalog no. Taxon LP1 WP1 LP2 WP2
6.3
7.4
LP3
2.5
3.2
WP3
9.1
12.9
12.4
LeP4
9.6
13.6
12.8
LiP4
5.0
6.6
7.0
WaP4
4.1
5.5
4.8
WblP4
4.6
7.7
6.5
LpP4
3.5
4.9
5.1
LmP4
9.3 10.6 10.7 8.2
10.5
LM1
13.3 14.4 14.9 9.1
11.5
WM1
6.8
6.4
LM2
11.7
9.0
WM2
192 L. Werdelin and R. Dehghani
LAET 78-5346 EP 818/98 EP 561/00 EP 270/000 EP 042/04 EP 890/05 LAET 75-2368 LAET 75-1959 LAET 74-289 EP 1169/05 LAET 77-4570 LAET 75-2722 LAET 78-4691 LAET 78-4955a, b LAET 75-3340 LAET 78-5298a, b LAET 75-2807a, b LAET 78-5295 LAET 78-4980 LAET 78-4736 LAET 75-3616 LIT 59/359 LAET 75-3565 LAET 75-940 LAET 75-2997 LAET 75-3368a LAET 76-3973 EP 015/98 EP 436/98 EP 1500/98 EP 466/00 EP 531/00 EP 1787/00 EP 1790/00 EP 1874/00 EP 2577/00 EP 2887/00 EP 2888/00 EP 4167/00 EP 4168/00 EP 035/01
Herpestes palaeoserengetensis Herpestes palaeoserengetensis Herpestes palaeoserengetensis Herpestes palaeoserengetensis Herpestes palaeoserengetensis Herpestes palaeoserengetensis Herpestes ichneumon Herpestes ichneumon Herpestes ichneumon Herpestes ichneumon Galerella sp. Galerella sp. Galerella sp. Galerella sp. Galerella sp. Helogale palaeogracilis Helogale palaeogracilis Helogale palaeogracilis Helogale palaeogracilis Helogale palaeogracilis Helogale palaeogracilis Helogale palaeogracilis Helogale palaeogracilis Helogale palaeogracilis Helogale palaeogracilis Helogale palaeogracilis Helogale palaeogracilis Helogale palaeogracilis Helogale palaeogracilis Helogale palaeogracilis Helogale palaeogracilis Helogale palaeogracilis Helogale palaeogracilis Helogale palaeogracilis Helogale palaeogracilis Helogale palaeogracilis Helogale palaeogracilis Helogale palaeogracilis Helogale palaeogracilis Helogale palaeogracilis Helogale palaeogracilis
2.2
1.9
2.5 2.7 2.0 2.0 2.1 1.9
1.3 1.3
1.4 1.2 1.4 1.5
1.8
1.6
4.4
4.0
4.9 4.4 3.9 3.8 3.9 3.6
2.6 2.6
2.7 2.7 1.9
2.8
2.4
3.1 2.9
1.6
1.7
3.1
2.1 1.7
1.7
1.7
2.9
3.7 2.9
3.3
2.9
1.7
1.4 1.6
2.8 3.0
1.9 2.1 2.0 2.0 2.2 2.3
3.4
1.8 2.0 2.0
2.1 2.1
2.1
3.8 3.7 3.3 3.4 3.5
3.8 3.9 3.6
3.5 3.6
3.7
3.3 4.1
3.7
2.1 2.1 2.0 2.2
2.0 2.0 2.1
1.6 1.5 1.6 1.5
3.9 3.4 3.7
2.7
2.2
3.6 3.0 2.6
2.5
4.1
4.4 4.5
5.3
7.3
4.5
2.8 2.6 2.6
Wp4
3.2 2.5 2.2 2.2
2.0
3.3
5.9 5.4 4.8
Lp4
6.3
2.1 2.3
Wp3
4.9 4.7
Table 8.3 Measurements of the lower dentition of Herpestidae, Viverridae, and Mustelidae from Laetoli Catalog number Taxon Lp2 Wp2 Lp3
3.8 4.0
3.8
5.1
4.5 4.7
4.5
4.4 4.1 4.5
4.3
6.2
8.1 7.5 5.9
7.6
6.2 5.3
Lm1
(continued)
2.5 2.6
2.5
2.4
2.8 2.8
2.7
2.5 2.3 2.7
3.4 2.4 2.1
4.5 4.2 3.5
3.6 2.9
Wm1
8 Carnivora 193
Table 8.3 (continued) Catalog number Taxon EP 348/01 Helogale palaeogracilis EP 467/01 Helogale palaeogracilis EP 636/01 Helogale palaeogracilis EP 642/01 Helogale palaeogracilis EP 390/03 Helogale palaeogracilis EP 770/03 Helogale palaeogracilis EP 873/03 Helogale palaeogracilis EP 985/03 Helogale palaeogracilis EP 2430/03 Helogale palaeogracilis EP 041/04 Helogale palaeogracilis EP 993/04 Helogale palaeogracilis EP 1456/04 Helogale palaeogracilis EP 1709/04 Helogale palaeogracilis EP 118/05 Helogale palaeogracilis EP 1224/05 Helogale palaeogracilis EP 3858a/00 Helogale palaeogracilis EP 3858b/00 Helogale palaeogracilis EP 1097/05 Helogale palaeogracilis LAET 75-3334 Helogale cf. palaeogracilis LAET 75-1974 Helogale cf. palaeogracilis EP 1324/04 Helogale cf. palaeogracilis LAET 75-3741 Mungos dietrichi LAET 75-2769 Mungos dietrichi LAET 75-3072 Mungos dietrichi LAET 77-4571 Mungos dietrichi EP 1217/03 Mungos dietrichi LAET 75-1923 Mungos sp. nov.? EP 544/01 Mungos sp. nov.? LAET 78-5315 Genetta sp. EP 389/03 Genetta sp. LAET 75-2661 aff. Viverridae EP 1140/01 Propoecilogale bolti EP 523/04 Mustelidae indet. For explanation of abbreviations see Material and Methods 4.4 4.5
4.7
1.9 1.4 1.5
1.3
2.3
1.8
3.2 2.9 2.4
2.8
3.8
2.8 2.6
1.6 1.6 1.7
1.8
2.8 2.8
1.4
1.8 1.8 1.8 1.8 1.7 1.6 1.7 2.3 1.6 1.4 1.5 1.5
2.9 3.9 3.2 3.2 3.1 3.0 3.3 3.6 3.0
2.7 2.8 2.9
Wp3 2.1 1.7
Lp3 3.2 2.6
Wp2
Lp2
5.5 4.9 6.0
4.2
3.3 3.5
3.6 3.6 3.2 3.6 4.1
2.2
1.7 2.3
2.2 2.3 1.9 2.3 2.1
1.8
3.4
3.6 3.6 3.7 3.7 3.6
Wp4 2.2
Lp4 3.3
5.1 6.2
5.6 5.3 5.7 5.6 5.2 6.9
4.5
3.9
2.1 3.0
4.2 4.0 4.0
4.0 3.6
2.6
2.6
2.7 2.5 2.7
2.4
4.0
5.0 4.5 4.2
Wm1 2.7
Lm1 4.0
194 L. Werdelin and R. Dehghani
LAET 78-5435a, b Herpestes palaeoserengetensis LAET 78-5286 Herpestes palaeoserengetensis LAET 76-3235 Herpestes palaeoserengetensis EP 270/00 Herpestes palaeoserengetensis LAET 75-2807a, b Helogale palaeogracilis LAET 75-2994 Helogale palaeogracilis LAET 75-2503 Helogale palaeogracilis LAET 75-3565 Helogale palaeogracilis LAET 75-940 Helogale palaeogracilis LAET 75-3741 Mungos dietrichi For explanation of abbreviations see Material and Methods
3.5 3.9 3.4 2.2 2.5
2.5
5.2 5.1 5.2 3.1 3.5
3.2
3.9 5.9
3.7
4.8 4.6 5.0
5.3 4.9 5.0 4.8 3.4
7.6 7.5 7.6 7.6 4.4
Table 8.4 Measurements of the upper dentition and cranium of Herpestidae, Viverridae, and Mustelidae from Laetoli Catalog number Taxon LP3 WP3 LP4 WP4
9.4
15.1
C-C
P-P
16.7
25.2
IOB
10.9
16.4
POC
10.0
13.3
ZB 44.5
8 Carnivora 195
Crocuta dietrichi Crocuta dietrichi Crocuta dietrichi Crocuta dietrichi Crocuta dietrichi Crocuta dietrichi Crocuta dietrichi Crocuta dietrichi Parahyaena howelli Parahyaena howelli Parahyaena howelli Parahyaena howelli Ikelohyaena cf. I. abronia Ikelohyaena cf. I. abronia 13.8 13.3 14.0
For explanation of abbreviations see Material and Methods
LAET 76-3970 LAET 74-149 LAET 75-2953 LAET 76-3951 LAET 78-5107 LAET 77-5370 EP 1067/04 EP 1390/05 EP 395/98 EP 463/01 EP 829/00 KK 82-58 LAET 75-3338 LAET 75-1849 10.7 12.3
Table 8.5 Measurements of the lower dentition of Laetoli Hyaenidae Catalog no. Species Lci Wci
8.9 9.9 7.7 6.5
8.7
12.6
15.0 14.5 13.4 12.1
8.0 8.6
Wp2
12.2 14.4
Lp2
12.8 10.7 7.9
12.5
19.1 19.7 18.5
12.5 11.3 12.6 11.0 12.3 12.3 11.5
Wp3
17.9 17.1 17.8 17.6 17.9 17.7 17.6
Lp3
15.4
20.4
20.7 19.8 19.8
Lp4
10.7 8.5 8.7
12.4 11.5 12.5
12.0
Wp4
10.2
11.2 10.4
10.9
Lpp4
23.7
23.4
26.6
Lm1
9.5 8.0
10.0
12.2
Wm1
19.4 13.9
20.7
22.7
Ltm1
196 L. Werdelin and R. Dehghani
LAET 74-185 Crocuta dietrichi EP 1067/04 Crocuta dietrichi LAET 76-4092 Parahyaena howelli LAET 76-4008 Parahyaena howelli KK 82-58 Parahyaena howelli LIT 59/465 Ikelohyaena cf. I. abronia EP 1046/98 Ikelohyaena cf. I. abronia EP 1218/03 Ikelohyaena cf. I. abronia 11.4 LAET 75-494 Lycyaenops cf. L. silberbergi For explanation of abbreviations see Material and Methods 7.9
Table 8.6 Measurements of the upper dentition of Laetoli Hyaenidae Catalog no. Species LCs WCs LP1 5.3
7.8 8.8 9.1 7.7 11.0
12.6 15.2 14.6 12.2 19.1
WP2 9.5
LP2 14.1
20.4 15.9 20.6
12.8 9.3 13.2
14.1
20.7
WP3 12.9
LP3 18.9
29.9 30.2
LP4
WaP4
15.7 15.7 13.5 18.7
15.9
WblP4
10.6
9.8 9.1
9.9
LpP4
13.1
10.1 11.1
11.4
12.9 11.3
LmP4
8 Carnivora 197
198
L. Werdelin and R. Dehghani
Table 8.7 Measurements of the lower dentition of Laetoli Felidae Catalog no. Taxon Lp3 Wp3 EP 1333/98 Panthera sp. aff. P. leo LAET 78-5122 Panthera sp. cf. P. pardus 12.8 LAET 75-537 Panthera sp. cf. P. pardus 10.9 EP 065/99 Panthera sp. cf. P. pardus 11.5 EP 1622/00 Panthera sp. cf. P. pardus EP 1621/00 Acinonyx sp. EP 927/01 Acinonyx sp. 13.0 LAET 75-991 Caracal sp. or Leptailurus sp. 6.5 EP 093/04 Caracal sp. or Leptailurus sp. EP 158/00 Caracal sp. or Leptailurus sp. 7.3 EP 3934/00 Caracal sp. or Leptailurus sp. 6.6 EP 119/01 Felis sp. 5.8 EP 120/01 Felis sp. For explanation of abbreviations see Material and Methods
7.2 5.6 5.9
7.1 3.3
Lp4
Wp4
17.1 14.8
8.6 7.1
16.6
8.6
15.8 8.4 8.1
8.0 3.8 3.9
Lpp4
7.8
4.5
Lm1
Wm1
Ltm1
27.6
13.1
27.0
16.3
7.6
18.9 14.4
8.5 6.6
10.2
4.3
13.6
4.1 8.7
9.9
2.3 6.8
3.3
Table 8.8 Measurements of the upper dentition of Laetoli Felidae Catalog no. Taxon LP3 WP3 LP4 WaP4 WM1 N’Garussi 1959
Panthera sp. aff. P. leo
24.9
13.2
35.5
19.9
12.7
For explanation of abbreviations see Material and Methods
Genus Nyctereutes Temminck, 1838 The genus Nyctereutes, raccoon dogs, is relatively well known in the Plio-Pleistocene of Eurasia, with a number of identified species (e.g., Tedford and Qiu 1991). In Africa the genus, which is absent from the modern fauna, is much more rare, with the only occurrences apart from the present one being from Ahl al Oughlam in Morocco and Kromdraai and Elandsfontein in South Africa (Ficcarelli et al. 1984; Geraads 1997).
?Nyctereutes barryi sp. nov. (Fig. 8.1) Holotype: LAET 75-3522 (Fig. 8.1; Barry 1987, fig. 7.8 [a, b as LAET 75-3562b]), mandibles and cranial fragment with left i2–m1, m2 alveolus, right i2–p1, p4 alveolus, m1–m2, left I1–C, P1 and P2 alveoli, P3–M1, right M1, right manus (Laetolil Beds, upper unit, Loc. 10W, between Tuff 3 and 3 m below Tuff 1). Synonymy: cf. Canis brevirostris Barry, 1987 Diagnosis: Medium-sized Canidae; premolars set close together, with no diastemata between them; M1 nearly rectangular in occlusal view; paracone and metacone of M1 separated from buccal margin of tooth by a stylar shelf; paraand metacristae of M1 strongly developed; hypocone smaller than protocone and crest-like; postprotocrista strongly developed; m2 relatively large; m1 and m2 nearly equal in width;
Fig. 8.1 ?Nyctereutes barryi, LAET 75-3522 (holotype). (a) Cranial fragment in ventral view. (b–d) Right mandibular ramus (labeled LAET 75-3562) in (b) occlusal, (c) buccal, and (d) lingual view
transverse crest uniting m1 hypoconid and entoconid lacking; small paraconid present on m2. Derivation of name: After Dr. John Barry in recognition of his major contributions to the study of African carnivores.
8 Carnivora
Additional specimens: Laetolil Beds, upper unit: LAET 74-249, right mandible fragment with p4–m1, left m1; LAET 76-3844, right mandible fragment with p3 and m1; LAET 75-3108, left edentulous mandible fragment; LAET 78-5251, mandible fragment with c, alveoli for p1–m2; LAET 78-5385, right mandible fragment with roots of p1–p3, broken p4, alveoli of m1–m3; EP 227/05, left mandible fragment with posterior root of p3, p4–m1; EP 597/01, left mandible fragment with m1; EP 1319/98, isolated right m1; ?EP 1765/03, left mandible fragment with roots of p2 and p3, anterior root of p4; EP 286/05, left mandible fragment with broken p2 and p3; EP 892/05, isolated left p2 or p3; ?EP 1049/98, broken right M1; EP 2238/00, left m1 talonid fragment. Description: This composite description will for the most part be based on LAET 75-3522, just as was the case in Barry (1987). Despite the recovery of a number of new specimens referable to this species, this specimen remains by far the best preserved. In fact, the cranial fragment and upper dentition of this specimen are still the only definite examples of these features, as EP 1049/98 can only tentatively be assigned to this species. The preserved incisors, I2 and I3, are simple, conical teeth with no caniniform development. Neither has a lingual accessory cusp, but I3 has a small cingulum on its mediolingual corner. The canine, although missing the tip, is short and relatively straight, with distinct mesial and distal crests. There is a short (ca. 4 mm) diastema between I3 and C. The alveolus for P1 is large and set directly behind the canine. It is followed immediately by the two alveoli for P2, of which the latter is damaged, so their relative sizes cannot be determined accurately. Both of these teeth must have been relatively large. The P3 appears to be of about the same size as P2. It is a simple tooth, lacking mesial and distal accessory cusps. The main cusp is set slightly mesial to the midline, between the mesial and distal roots. The posterior margin of the cusp is somewhat crestlike and distinctly concave. The distal end of the tooth is slightly elongated. The P4 is robust. The mesiobuccal corner is large and rounded. The paracone is tall, with a straight mesial margin. The preparacrista extends to the buccal margin of the small protocone. The postparacrista leads to a shallow carnassial notch, followed by an elongated metastyle. There is a small lingual cingulum at the base of the metastyle. The M1 is the most distinctive tooth in the upper jaw. The paracone and metacone are both low and separated from the buccal margin of the tooth by a stylar shelf bounded buccally by a strong cingulum. The para- and metacristae are very strongly developed, forming a nearly continuous crest from the mesiobuccal corner of the tooth to the distal-most end. A similar development is seen in the pre- and postprotocristae. The protocone is set mesiolingually and is bordered mesially by a strong cingulum. The hypocone is smaller than the protocone and set distally about halfway between the buccal and lingual margins. Like the other cusps, the hypocone has strong
199
pre- and postcristae. Thus, the center of the tooth between the cusps is surrounded by a nearly continuous crest. Distal to the protocone and lingual to the hypocone there is a broad, flat shelf. Overall, the tooth is relatively rectangular in occlusal view. Barry (1987) describes and figures a right M2 in this specimen. This tooth has subsequently been lost. He describes it as similar in general appearance to M1, but relatively broader and with a less developed buccal cingulum. The cranium of this specimen is too damaged to provide any useful information on morphological structures, except to note that the anterior margin of the orbit lies above the anterior half of P4. The lower dentition, with the exception of i2 and m3, is known in its entirety from LAET 75-3522. A number of other specimens, as listed above, provide information on variation. The mandibular ramus is slender and low with a very weakly developed subangular lobe. The symphyseal rugosity is nearly horizontal and extends distally to the gap between p2 and p3. The masseteric fossa is deep, tall, and short. It extends mesially to about the posterior end of m3. There are two mental foramina: a larger one situated beneath p1 and p2 and a smaller one situated beneath the posterior root of p3. The cheek teeth are set close together with no or only very small diastemata between them. The incisors are relatively procumbent, though this may in part be a postmortem effect. Both i2 and i3 are similar in general morphology. They are somewhat spatulate and lack distinct medial and lateral accessory cusps. The canine is short, robust, and relatively straight. It has distinct mesial and distal crests. The p1 is a small, single-rooted tooth with a triangular main cusp that is set slightly anterior to the middle of the tooth. There are no mesial and distal accessory cusps, but there is a very small distal bump indicating an incipient cingulum. The p2 is similar to p1, but it is larger and two rooted. The p3 is similar to p2 and only slightly larger. The p4 is slightly larger than p3 and has a distinct, crestlike distal accessory cusp situated at the middle of the distal face of the main cusp. Posterior to this accessory cusp there is a small cingular cusp. The lower carnassial is stout. The trigonid constitutes about 3/5 of the total length of the tooth. The paraconid is short and low, and the protoconid is longer and taller. The postprotocristid is strongly developed and meets the metacristid at a shallow notch. The metaconid is low and blunt yet relatively large for a canid. It is about equal in height to the well-developed hypoconid. The latter is connected to the trigonid through a strong cristid obliqua, which runs mesially and slightly lingually from the hypoconid and ends at a shallow postvallid notch. The entoconid is considerably lower and smaller than the hypoconid. There is no transverse crest connecting the two talonid cusps. The m2 is large and nearly equal in width to the m1. There is a small paraconid at the mesial end of the tooth. Mesiobuccal to this, there is a short cingulum shelf. The paraconid and
200
L. Werdelin and R. Dehghani
Fig. 8.2 Bivariate diagram of lengths of P4 and M1 in selected Canidae. The relative lengths of these teeth are similar in LAET 75-3522 and N. donnezani, a raccoon-dog from the Pliocene of Europe. The
proportions of C. adustus are also similar, but the morphology is different. EP 437a/98 has a much shorter M1, similar to the situation in C. aureus and C. mesomelas
protoconid are connected by crests that are interrupted by a shallow notch. The protoconid is robust and triangular in occlusal view. The metaconid is similar in size to the protoconid. The hypoconid is large and connected to the talonid by a distinct prehypocristid. It is set nearly at the distobuccal corner of the tooth. The entoconid is composed of two small cuspids on the lingual margin of the tooth. It is followed distally by a low ridge that runs to the distal-most end of the tooth. The m3 is not present but was a small, single-rooted tooth. Discussion: This taxon was extensively discussed by Barry (1987). It displays several features of the lower dentition that differentiate it from the Canis lineage, including the absence of a transverse crest uniting the hypoconid and entoconid on m1, the closely set premolars, and the relatively broad m1. However, it is in the features of the upper dentition that the differences are clearest. ?Nyctereutes barryi has a relatively wide P4 with a very small, anteriorly placed protocone and an M1 that is very wide lingually, with a prominent protocone-hypocone crest. Other features that differentiate it from Canis are listed in Barry (1987: 240). In a number of features, ?N. barryi resembles primitive species of Nyctereutes. This is especially true of the upper dentition, which is quite similar in some respects to the primitive N. tingi from China (Tedford and Qiu 1991), though the molars of the Laetoli form appear more derived, especially in the subequal paracone and metacone, and are in some respects similar in morphology to the more derived N. sinensis (Tedford and Qiu 1991; Fig. 8.2). Like N. tingii, ?N. barryi lacks the subangular lobe of the mandibular ramus characteristic of
more derived species of Nyctereutes. Barry (1987) also records similarities between the Laetoli form and the European N. donnezani (Soria and Aguirre 1976), while noting that the latter is more derived in several features. N. donnezani is also more derived than N. tingi, and, if the generic attribution is valid, which remains somewhat uncertain, ?N. barryi is morphologically intermediate between N. tingi and N. donnezani in the raccoon-dog lineage.
Genus ?Canis The earliest definite record of Canis from Africa, or anywhere, is from South Turkwel at ca. 3.5 Ma (Werdelin and Lewis 2000). Thus, if the Laetoli material is demonstrated to belong to Canis, it could be the oldest of the genus on record. The later record of Canis in Africa is persistent but spotty, with limited material being found at many sites from the late Pliocene onwards.
cf. Canis sp. A (Fig. 8.3) Specimens: Laetolil Beds, upper unit: LAET 77-4603, left mandible fragment with p3–p4, roots of p2, and anterior root of m1; EP 1047/98, left maxilla fragment with I2–C roots, P1–P2, P3 anterior alveolus; EP 2431/03, left mandible fragment with p4; EP 437a/98, right maxilla with P4–M1; EP 437c/98, left maxilla with M1–M2 (Fig. 8.3); EP 437b/98, right maxilla with P2–M2, cranial fragments.
8 Carnivora
Description: The mandibular ramus is quite robust and deep, with at least two mental foramina. The larger of the two is situated beneath p1 and the smaller beneath p2. The cheek teeth are set well apart, with distinct diastemata between them. Both p2 and p3 are relatively slender. The p4 appears more robust. The main cusp is set slightly mesial to the middle of the tooth. It is fairly tall, and its distal face has a low but long accessory cusp, followed by a distinct posterior cusplet and cingulum. The P4 is short, with a small, low protocone, a tall paracone, and a short, stout metastyle.
Fig. 8.3 cf. Canis sp. A, EP 437c/98. Left maxilla fragment in occlusal view
Fig. 8.4 Bivariate diagram of length and width of M1 in selected Canidae. The three specimens numbered EP 437/98 all have a shorter and relatively wider M1 than in modern Canis spp., while the M1 of
201
The M1 is triangular rather than rectangular in occlusal view. The length/width relationship and its differences from ?N. barryi can be seen in Fig. 8.4. It has a low, blunt paracone, which is bounded mesiobuccally by a broad parastyle shelf. The metacone is also low and blunt. On its buccal side there is a broad metastyle shelf, though it is not as broad as the parastyle shelf. These two shelves may (EP 437a/98) or may not (EP 437c/98) be connected. None of the available specimens show the features of the protocone clearly, but it was seemingly robust. The M2 is similar in structure to M1, but smaller and relatively mesiodistally shorter; its metastyle shelf is smaller than in M1 and is connected to the parastyle shelf. Discussion: These specimens are united in being roughly the size of a medium-large jackal; in having premolars that are widely spaced as in Canis, in contrast to the previous taxon; and in having upper molars that are quite different from those of ?N. barryi and, again, more Canis-like. The p4 of LAET 77-4603 and EP 2431/03 are the size of a large jackal p4. The P4 and upper molars of the three specimens collected under the catalog number EP 437/98, on the other hand, are somewhat smaller (Fig. 8.4), and, if these specimens all belong to the same taxon, it is likely that there is a difference between it and modern jackals in the relative sizes of the premolars and molars. There are clear differences in a number of characters between this material and ?N. barryi, and, although the fragmentary nature of the material makes it impossible at present to definitely assign it to Canis, the few characters available suggest that that is the most plausible generic allocation. There are
LAET 75-3522 is relatively narrower. The much smaller EP 1245/01 is here referred to aff. Otocyon sp.
202
similarities in size and morphology to undescribed material of Canidae from Hadar (Werdelin and Lewis, unpublished), and a detailed comparison with this material may allow for a more specific determination than is presently possible. cf. Canis sp. B (Fig. 8.5) Specimens: Laetolil Beds, upper unit: LAET 78-4713 (Fig. 8.5; Barry 1987, fig. 7.7), isolated left m1 talonid; EP 2005/00, isolated right m1 talonid. Description: Both specimens are broken posterior parts of lower carnassials. LAET 78-4713 preserves a larger part of the tooth but is more damaged than EP 2005/00. The protoconid is robust. As is normal in canids, the metaconid is quite low and set slightly distal to the protoconid. The hypoconid is large and has a cristid obliqua that runs mesially to the postvallid notch. The entoconid is very low, hardly more than a bump on the lingual margin of the tooth. It is connected to the hypoconid by a low crest that includes a blunt hypoconulid. Discussion: Barry (1987) provides an extensive discussion of this taxon based on specimen LAET 78-4713. Specimen EP 2005/00 is nearly identical to LAET 78-4713 but is slightly more worn and preserved quite differently. Barry’s (1987) discussion concerns the generic identity of the specimen, based on his belief that the morphology of the talonid, which lacks a true cristid obliqua and transverse crest uniting the hypoconid and entoconid (Barry 1987: 236), precludes allocation to Canis. However, canids in which the hypoconid is strongly developed and the entoconid weak seem often to have a cristid obliqua that is directed mesially rather than mesiolingually. Such is the case with the Canis sp. from South Turkwel (specimen KNM-ST 22822; Werdelin and Lewis 2000), a site that is roughly contemporaneous with the Upper Laetolil Beds. Since this is the oldest known Canis, as attested to by the transverse crest linking the hypoconid and entoconid, the mesially oriented cristid obliqua may perhaps be the primitive condition in this genus. The absence of the transverse crest is a more serious impediment to the assignment of LAET 78-4713 and EP 2005/00 to Canis. However, although an entoconid is present in both specimens, it is very low and is more of a distolingual ridge than a proper cusp. This reduction of the entoconid
Fig. 8.5 cf. Canis sp. B, LAET 78-4713. Left m1 protoconid and talonid in (a) occlusal, (b) buccal, and (c) lingual view
L. Werdelin and R. Dehghani
may also have obliterated the transverse crest. Thus, we find the arguments against placing these specimens in Canis weakened though still valid. For the time being, we prefer to place them in cf. Canis sp. to indicate our opinion of their probable affinities. Barry (1987) suggests that they belong to a form much larger than that represented by the previous taxon. We believe that this size difference is exaggerated, and, although it seems unlikely that Canis sp. A and Canis sp. B are the same taxon, we cannot definitively rule this out.
Genus Otocyon Müller, 1836 The genus Otocyon appears late in the African fossil record, with its only appearance at Lainyamok in Kenya at ca. 300 ka (Potts and Deino 1995). An older record from Olduvai Bed I is referred to the doubtfully distinct genus Prototocyon (Petter 1973). aff. Otocyon sp. (Fig. 8.6) Specimens: Laetolil Beds, upper unit: LAET 75-1419 (Petter 1987: Plate 2, fig. 15), right mandible fragment with alveoli for p1–m3; LAET 75-2812, right distal tibia; LAET 75-3814, right radius lacking proximal end, right distal ulna fragment; LAET 76-3936 (Petter 1987: Plate 1, fig. 9), isolated left m2 or m3; EP 2126/00, right mandible fragment with root of c, p1–p4 (Fig. 8.6); EP 208/01, left maxilla fragment with C, separate fragment with P3–P4; EP 1245/01, isolated right M1; EP 1630/98, right dp4. Description: The mandibular ramus is very slender. There are at least two mental foramina. The anterior, larger one lies beneath p1, while the posterior, smaller one lies beneath
Fig. 8.6 aff. Otocyon sp., EP 2126/00. Right mandibular ramus in (a) occlusal, (b) buccal, and (c) lingual view
8 Carnivora
the posterior end of p3. In EP 2126/00, the latter foramen is double, with the posterior of the pair lying beneath the gap between p3 and p4. The symphysis is short, extending only as far distally as p2. The p1 is a small, single-rooted tooth that is separated from the lower canine by a short diastema. The p2 is a small tooth, triangular in lateral view, that is separated from p1 and p3 by gaps of 1–2 mm. The main and only cusp of p2 is set nearly at the middle of the tooth. The distal end of the tooth is slightly wider than the mesial end. The p3 is similar to p2 but has a small swelling on the distal edge of the main cusp and a distinct distal cusp at the base of the crown. The p4 is similar to the p2 and p3, but the swelling on the distal edge of the main cusp has here developed into a distinct, ridge-like accessory cusp. Distal to this is a cusp that lies at the base of the crown. The P3 is a slender, triangular tooth lacking mesial and distal accessory cusps but with a small cusp at the distal base of the crown. The P4 is short and relatively robust. The parastyle is small, whereas the protocone is relatively large and strong. There is a strong invagination in the middle of the mesial margin of the tooth. The paracone is tall and robust, and the metastyle is short and stout. There is a weak cingulum at the lingual side of the metastyle. The M1 is more or less rectangular in occlusal view, though the metastyle wing extends farther distally than the rest of the distal margin. The paracone is large and blunt. It is separated from the metacone by a distinct notch. The protocone is broad but low, and the center of the tooth is damaged and features are difficult to discern. The stylar shelf is narrow mesially and buccally but broadens considerably distal to the metacone.
203
Discussion: This material shows the presence at Laetoli of at least one species of small canid the size of a fox. All of the material is smaller than the homologous elements in jackals. The mandible fragment EP 2126/00 is slightly smaller than a jackal and has relatively short and widely spaced premolars, whereas the maxilla fragment EP 208/01 and the M1 EP 1245/01 are considerably smaller than a jackal (Fig. 8.7). These may represent different taxa or simply the sexes of a single species. The morphology of the lower premolars is strongly reminiscent of that of the bateared fox, Otocyon. The earliest currently known record of this lineage is Prototocyon recki from Olduvai Gorge Bed I (Pohle 1928; Petter 1973). At present, there is no reason to believe that more than one taxon is involved, and the Laetoli specimens may represent the earliest known member of the Otocyon lineage, which is currently considered the sister group to all other Vulpini (Lindblad-Toh et al. 2005).
Family Mustelidae Fischer, 1817 The fossil record of Mustelidae in Africa is uneven. Some Lutrinae are well represented, especially the lineage of giant bunodont forms referred to the Enhydrini (Morales and Pickford 2005). The Mellivora lineage is sporadically represented from the Late Miocene onwards. However, small Mustelidae, especially of the Plio-Pleistocene, are rare, being confined to a handful of records (Werdelin and Lewis 2005). Metric data for Laetoli Mustelidae are given in Tables 8.3 and 8.4.
Fig. 8.7 Bivariate diagram of length and width of p2 in selected Canidae. The p2 of EP 2126/00 (aff. Otocyon sp.) is shorter and broader than in Canis spp. or ?N. barryi
204
Genus Prepoecilogale Petter, 1987 This genus is known only from Laetoli and from Bolt’s Farm in South Africa. It is closely related to Poecilogale and Ictonyx (Petter 1987).
Prepoecilogale bolti (Cooke, 1985) (Fig. 8.8) Specimens: Laetolil Beds, upper unit: LAET 75-1358 (Fig. 8.8; Petter 1987: Plate 1, fig. 3), left maxilla fragment with P3–P4, left premaxilla fragment with incisor, left mandible fragment with p4-m1, left calcaneum, skull fragment, vertebral fragments; LAET 74-248 (Petter 1987: Plate 1, figs. 4–7), maxilla fragment with C, P2–P3, left and right tympanic bullae, right humerus, distal left humerus, proximal right ulna, proximal right femur, distal right femur, proximal and distal right tibia, innominate fragment, left and right astragali, vertebral fragments; EP 634/03, left P4 metastyle; EP 466/01, left calcaneum. Ndolanya Beds, upper unit: EP 789/01, left calcaneum, EP 1140/01, right mandible fragment with m1–m2. Description: The upper canine is robust and has an oval, nearly round cross-section. It is slightly curved. There is a diastema of just less than 1 mm to the P2, which is small and double-rooted. There is no mesial accessory cusp, and the main cusp is situated nearly at the mesial extremity of the tooth. The posterior shelf is long but lacks a cusp. There is a weak basal cingulum on the lingual side of the tooth. The P3 is a short tooth with small mesial and distal basal cusps. The main cusp is short, nearly symmetrical, and very tall and lies just anterior to the middle of the tooth. The tooth is surrounded by a moderate to weak cingulum. The P4 is slender. The protocone extends far mesial to the parastyle.
L. Werdelin and R. Dehghani
The paracone is tall and trenchant and continues to the metastyle without a carnassial notch, as is normal in Mustelidae. The p4 is similar to P3 in being short and with a very tall, nearly symmetrical main cusp. This cusp has a very small accessory cusp on its distal face. There are low mesial and distal basal cusps and a strong buccal cingulum. The m1 has a short, low paraconid and a taller, longer protoconid. The metaconid is small and located buccal to the distal end of the protoconid. The talonid is short, with a trenchant hypoconid but no entoconid. The m2 is very small and round in occlusal view. Discussion: This material all clearly represents a mustelid of very small size and is, as noted by Petter (1987), indistinguishable from “Ictonyx” bolti from Bolt’s Farm (Cooke 1985). Small mustelids are exceedingly rare in the African post-Miocene fossil record. Apart from the present material, the only record is from Hadar, where several undescribed taxa of small mustelids have been found.
Prepoecilogale sp. Specimens: Laetolil Beds, upper unit: EP 2889/00, broken left calcaneum. Ndolanya Beds, upper unit: EP 3324/00, right calcaneum. Description: The postcranial specimens are typical mustelid in morphology but are not sufficient to definitively identify them as P. bolti.
Genus Mellivora Storr, 1780 Mellivora probably evolved in the latest Miocene from Erokomellivora or a similar form (Werdelin 2003a). The question of the distinction, if any, between M. benfieldi from, for example, Langebaanweg (Hendey 1974) and the extant M. capensis (Petter 1987) cannot be resolved here.
Mellivora sp.
Fig. 8.8 Prepoecilogale bolti, LAET 75-1358. (a–c) Left maxilla fragment in (a) occlusal, (b) lingual, and (c) buccal view. (d–f) Left mandibular fragment in (d) lingual, (e) buccal, and (f) occlusal view
Specimens: Laetolil Beds, upper unit: LAET 75-530 (Petter 1987: Plate 1, fig. 12 [as LAET 531]; Plate 2, fig. 8), postcranial fragments, including ribs, axis, cervical, thoracic, and lumbar vertebrae, sacrum, proximal and distal phalanges; LAET 78-5078, proximal radius. Description: This material was described and discussed in detail by Petter, (1987), and this need not be repeated here as no new material of Mellivora has been recovered from Laetoli since then.
8 Carnivora
Tribe Aonyxini Sokolov, 1973 Unlike the Enhydrini, members of which are not present at Laetoli, Aonyxini is poorly represented in the African fossil record. Apart from the Laetoli record, the tribe is known only from some isolated elements from Ethiopia, Kenya, and South Africa. Aonyxini gen. et sp. nov. (Fig. 8.9) Specimens: Laetolil Beds, lower unit: KK 82-204, partial juvenile skeleton including maxilla fragment with C root, broken P3-M1, and posterior part of cranium (Fig. 8.9). Description: The skeleton includes many individual bones but is very poorly preserved. Because the individual is a juvenile, the only diagnostic element is the maxilla fragment, so this description will be restricted to that fragment. The canine is a rounded oval in occlusal view and lacks mesial and distal ridges or keels. The postcanine diastema is very short, no more than 2 mm. The P3 is very small and only slightly longer than it is wide. Its long axis is set somewhat at an angle to the tooth row. The P4 is rhomboid in occlusal view. The buccal side of the tooth is broken, but the outline remains, showing that the protocone shelf extends farther mesially than the paracone. The protocone itself is low. The distolingual shelf tapers gradually, and there is no distinct hypocone. The M1 has also been broken buccally. Its most prominent features are the strong marginal ridges on the lingual side of the tooth. The tooth is square in occlusal view, whereas other M1 of Lutrinae are, like the M1 of most Carnivora, wider than they are long. Discussion: These fragments certainly record the presence of a species of lutrine mustelid at Laetoli. However, the exact
Fig. 8.9 Aonyxini gen. et sp. nov., KK 82-204. (a) Posterior cranial fragment in ventral view. (b) Right maxilla fragment in occlusal view
205
relationships of the material remain obscure. The referral to Aonyxini is based on the combination of a long lingual shelf on the P4 and the absence of a hypocone. In Lutrini, the shelf normally extends only to the midpoint of the tooth or less, while in the Enhydrini a hypocone is present. However, the presence of only two upper premolars and the nearly square M1 are unique features within the Aonyxini (and, indeed, within the Lutrinae, as far as we are aware). At the very least, a new genus is indicated.
Mustelidae indet. Specimens: Laetolil Beds, upper unit: EP 523/04, left mandible fragment with broken m1, alveolus for m2. Description: The mandibular ramus is robust and tall, with a deeply excavated but in its preserved part dorsoventrally low masseteric fossa that extends approximately to the distal end of the m2 alveolus. The m1 is a slender tooth. The buccal trigonid cusps are set at a slight angle to the ramus. The presence or absence of a metaconid cannot be determined. The talonid is short, with a low, trenchant hypoconid and a small, distal entoconid. The m2 is single-rooted and slightly longer than it is wide. Discussion: This specimen is morphologically very similar to Mellivora sp. but is smaller than any individual of Mellivora known to us. Therefore, we have left it as Mustelidae indet. for the time being.
Family Viverridae Gray, 1821 Viverridae is better known in the fossil record than Herpestidae, though there is considerable confusion regarding the identity of early forms in this family. The Stenoplesictinae, with a fossil record extending back into the European Oligocene, is generally considered the oldest subfamily within the Viverridae (Hunt 1998), although if it associated with the Percrocutidae, as suggested by some authors (Morales et al. 1998; Morales et al. 2003), this would have to be reassessed. In Africa, the genus Herpestides, positively identified as being within the Viverridae by Hunt (1991), is known from the early Miocene (Schmidt-Kittler 1987). Many members of the Viverridae from the fossil record of Africa are large forms, in many cases considerably larger than any viverrid alive today [e.g., Hunt (1996)]. Smaller viverrids have a much more restricted fossil record. In eastern Africa, smaller viverrids are known from Lothagam, Middle Awash Adu-Asa Fm., Kanapoi, Allia Bay, Omo Shungura Fm. Mb. B, and the Upper Burgi and Okote Mbs. at Koobi Fora, in addition to Laetoli (HaileSelassie 2001; Werdelin 2003a, b). Metric data for Laetoli Viverridae are given in Tables 8.3 and 8.4.
206
L. Werdelin and R. Dehghani
Genus Viverra Linnaeus, 1758 Viverra is not well represented in the African fossil record. Two taxa are known, the Late Miocene V. howelli and the Late Miocene–Early Pliocene V. leakeyi, the latter of which is known from several localities in southern and eastern Africa (Petter 1963; Rook and Martinéz-Navarro 2004). Some authors place the latter species in the genus Megaviverra (Morales et al. 2005), but we prefer to keep it in the genus Viverra for reasons discussed elsewhere (Werdelin 2003a).
Viverra leakeyi Petter, 1963 Specimens: Laetolil Beds, upper unit: LIT 59-466, isolated C, P4, M1, M2 (holotype); LAET 75-2725 (Petter 1987: Plate 2, fig. 5), mandible fragment with p1. Description: This material was described and discussed in detail by Petter (1963, 1987), and this need not be repeated here, as no new material of this taxon has been recovered from Laetoli.
Genus Genetta Cuvier, 1817 Genetta sp. has been recorded from a number of Miocene localities in Africa (e.g., Beni Mellal [Ginsburg 1977]), but in all these cases the referral to Genetta is doubtful. The oldest certain record is from Kanapoi, of a species close to the extant G. genetta (Werdelin 2003b). The latter is known from the early Pleistocene (Werdelin and Lewis 2005).
Genetta sp. (Fig. 8.10) Specimens: Laetolil Beds, upper unit: LAET 78-5315, partial right mandible with m1 and alveolus for m2 (Fig. 8.10); EP 389/03, right mandible fragment with p1, broken p2, complete p3. Description: The mandibular ramus is deep relative to the size of the teeth. There are two mental foramina; the mesial is situated beneath the diastema between p1 and p2, and the distal is situated beneath the distal root of p3. The p1 is a small, single-rooted tooth. It is directed mesially and dorsally and has a distinct distal accessory cusp posterior to the buccolingually compressed main cusp. Although the p2 is broken mesially, it is clear that it is separated from the p1 by a diastema of approximately 1 mm. The p2 has a trenchant accessory cusp on the distal cingulum. The p3 is tall and buccolingually compressed. It has a prominent mesial accessory cusp; a very tall, trenchant main cusp; and a large distal accessory cusp that is appressed to
Fig. 8.10 Genetta sp., LAET 78-5315, right mandibular ramus in (a) buccal, (b) lingual, and (c) occlusal view
the distal face of the main cusp. Distal to this, the distal cingulum also bears a distinct cusp. The trigonid cusps of the carnassial are well developed and triangular in occlusal view, as is the trigonid itself. The protoconid is situated distobuccal to the paraconid, representing the buccal-most, and hence the widest, part of the tooth. Posterior to the paraconid and somewhat lingual to it is the metaconid. The buccal faces of the para- and protoconid are adapted for shearing, and their shearing blades meet at nearly right angles. The metaconid is lower and more conical than the other trigonid cusps. The talonid is much lower than the trigonid and forms a posterior ridge. The distobuccal portion of the talonid is damaged. Discussion: The lower carnassial is reduced in size relative to that of Herpestes. This reduction is also seen in the modern genet, G. genetta, which, together with tooth morphology, supports generic attribution of LAET 78-5315 to Genetta. In addition, the m1 of this specimen and specimens of H. ichneumon are similar in morphology, but they differ in ramus morphology. The horizontal ramus of LAET 78-5315 is more slender and shallower below m1, as it is in modern genets, in comparison with the more massive ramus of H. ichneumon. The anterior premolars in EP 389/03 are comparable to those of modern genets and are likely to belong to the same taxon as LAET 78-5315. The size and shape of the lower carnassial of LAET 78-5315 are similar to the extant viverrid species G. genetta (Fig. 8.11). The metaconid, however, is lower relative to the two other trigonid cusps in G. genetta than in LAET 78-5315. aff. Viverridae Specimens: Laetolil Beds, upper unit: LAET 75-2661, partial right m1.
8 Carnivora
207
Description: This partial right m1 preserves the paraconid, the mesiobuccal part of the protoconid, and the mesial root. The paraconid and buccal part of the protoconid make a shearing facet. The protoconid is taller than the paraconid. Discussion: This specimen shows general viverrid characters, but the state of preservation does not allow for a more specific determination.
early Miocene, the relationships between these earlier forms, such as Kichechia and Legetetia, and extant herpestids are not known. Laetoli is the source of by far the largest sample of herpestids in eastern Africa and plays a key role in understanding the modern herpestid fauna. Metric data for Laetoli Herpestidae are given in Tables 8.3 and 8.4.
Family Herpestidae Bonaparte, 1845
Genus Herpestes Illiger, 1811
The Herpestidae is the least-studied carnivoran family in Africa. Although known from records extending back to the
Herpestes has been reported from the Adu-Asa Fm., Middle Awash Valley, Ethiopia, by Haile-Selassie (2001). If this referral is correct, these are the oldest Herpestes specimens known. The genus is otherwise rare in the fossil record, being recognized only in the Denen Dora Mb. of the Hadar Fm., Ethiopia, and in Bed I of Olduvai, Tanzania (L.W., personal observation). In the modern African fauna, Herpestes, as used herein, is restricted to the Egyptian mongoose or ichneumon, H. ichneumon, which has a broad distribution across sub-Saharan Africa, except in the arid regions of southern and southwestern Africa. In the north, it extends along the Nile Valley but does not occur elsewhere in northern Africa (Kingdon 1997).
Herpestes palaeoserengetensis Dietrich, 1942 (Fig. 8.12) Fig. 8.11 Bivariate diagram of length and width of m1 in selected small Carnivora. The m1 of LAET 78-5315 closely resembles the m1 of G. genetta in proportions
Fig. 8.12 Herpestes palaeoserengetensis. (a-d) LAET 78-5435; cranium in (a) ventral and (b) left lateral view; mandible fragment in (c) left lateral and (d) occlusal view. (e–g) EP 818/98; left mandibular ramus in (e) buccal, (f) lingual, and (g) occlusal view
Specimens: Laetolil Beds, upper unit: LAET 76-3235 (Petter 1987: Plate 7.1, fig. 8), partial left maxilla with P3–P4 and a lower tooth fragment in occlusion with P3 and P4;
208
LAET 78-5286, isolated left P4, mesiobuccal root broken; LAET 78-5346, partial left mandible with c, p2–p4 and alveolus for p1, all teeth more or less broken; LAET 78-5435 (Fig. 8.12a–d; Petter 1987: Plate 7.1, fig. 1), cranium (5435a) with right P1–M2 and left C–M2, and mandible fragments (5435b) with left p2–m1, p2 and m1 broken, and right p1–p3, p1 and p3 broken, and an additional mandible fragment with right m1–m2, both teeth in poor condition; EP 890/05, mandible fragments with left c, alveoli of p2–p4, right c–p2, alveoli of p3; EP 818/98 (Fig. 8.12e–g), mandible fragment with p3–m1; EP 561/00, left mandible and maxilla with P2–M1, c–m1 (still in occlusion), right mandible with broken p4–m2, left distal humerus, proximal ulna, radius lacking distal epiphysis, right distal humerus, left proximal humerus; EP 042/04, mandible fragment with p3–p4; EP 270/00, left and right C, left c, right P4 fragment, right mandible fragment with broken p4–m2 (in matrix), right maxilla fragment with P3–M1, right proximal tibia, three innominate fragments, right calcaneum, right astragalus, phalanx, podial, two tibia shaft fragments, distal tibia, left and right distal femora, petrosal, and bone fragments. Description: The cranium LAET 78-5435a is well preserved but slightly deformed and broken as a result of postmortem damage. The postorbital process is not closed, as is true of the majority of Herpestes ichneumon specimens, whereas the orbit is closed in most specimens of Galerella sanguinea. The frontal is slightly arched posterior to the postorbital constriction. The braincase is elongated and has its maximum width at the level of the posterior process of the zygomatic arch. The sagittal crest is not well marked or ridgelike, and the posterior part of the parietal slopes downward fairly steeply to the nuchal crest. The tympanic bullae are well developed and inflated. The ectotympanic is tubular and forms an incomplete circle. The entotympanic chamber is inflated and spherical in shape, forming the deepest part of the bullae. The inflation of the posterior chamber is oriented posteroventrally. The paroccipital is closely appressed to the posterior part of the entotympanic without extending into the paroccipital process. There is no distinct mastoid process. Skull width is intermediate between that of the larger Herpestes ichneumon and the smaller Galerella pulverulenta. The postorbital constriction is narrower in LAET 78-5435a than in either H. ichneumon or G. pulverulenta, although only marginally more so than in the latter species. The width of the skull between the buccal margins of P4 and the least width of the skull at the postorbital constriction, relative to the least width between the orbits, are greater in H. ichneumon and G. pulverulenta than in LAET 78-5435a. The P1 is single-rooted, whereas P2 has two roots. The apex of the cusp of this latter tooth is located slightly posterior to a position between the two roots. The P3 has three roots, mesial, distal, and lingual. The lingual root is smaller and supports a small cusplet at the middle of the lingual face.
L. Werdelin and R. Dehghani
There are small basal cingula anteriorly and posteriorly, the latter of which is somewhat ridgelike. The P4 has three roots. The parastyles of both carnassials are worn. The paracone is taller than the protocone. The parastyle is located slightly anterior to the protocone, and the paracone is situated somewhat more buccally than lingually. The metastyle blade is in distal contact with the anterobuccal corner of M1. The M1 is triangular in shape and is mesiodistally short and buccolingually wide. The protocone is the largest of the three cusps. The M2 is small but similar in morphology to the M1. The p1 is small. The p2 and p3 cusp apices are located between their respective roots. They are both two-rooted and have a posterior accessory cusp and basal cingula mesially and distally. The p4 has the same basic morphology as p2 and p3, but with a more developed posterior accessory cusp and stronger cingulum. The crowns of both m1 and m2 are damaged. An anterior mental foramen is located beneath the anterior edge of p2 and a more posterior one beneath the p2–p3 diastema. Discussion: The majority of this material was attributed to Galerella palaeoserengetensis by Petter (1987) because of the conformity of cranial size and morphology of the tympanic bullae between LAET 78-5435a and the type specimen of Herpestes palaeoserengetensis, MB Ma 29566 (Dietrich 1942; Petter 1987). The new material from the Eyasi Plateau project matches that material in every respect. Petter made the distinction between Herpestes and Galerella on basis of the length of the cranium. In addition, Petter (1973) had described the species Galerella primitivus from Bed I at Olduvai, but he later (Petter 1987) synonymized this species with G. palaeoserengetensis from Laetoli on the basis of the increased sample of Laetoli material and an expanded comparative base. The importance of bulla morphology to the generic attribution may be questioned, however. Hunt (1974) showed that the morphology of the auditory region in Carnivora is applicable to systematic problems at the family level but is less useful at lower levels. The dentition of the Laetoli specimens matches that of extant Herpestes from Africa most closely in size, whereas the teeth of Galerella are generally more reduced. Other features supporting an inclusion in Herpestes are the retained first premolar in the lower jaw in the Laetoli specimens, as in extant Herpestes, and total skull size. The skull of LAET 78-5435a measures almost 80 mm, despite the frontal breakage, in which feature it resembles larger species of Herpestes, whereas Galerella is usually smaller.
Herpestes cf. H. palaeoserengetensis Specimens: Laetolil Beds, upper unit: LAET 78-4677, partial right mandible with p1 and broken p2, canine root and anterior p3 alveolus.
8 Carnivora
Description: The p1 is small and narrow, with the apex anteriorly oriented and a small elevation distally. The p2 has two roots and the crown is broken. There is an anterior mental foramen below p1 and a posterior mental foramen, presumably below the anterior root of p3. Discussion: This partial mandible is comparable in size to that of Herpestes palaeoserengetensis as described above. Features including size and the retained first molar suggest that this specimen belongs in Herpestes palaeoserengetensis. Only the state of preservation precludes a positive identification.
Herpestes ichneumon Linnaeus, 1758 Specimens: Laetolil Beds, upper unit: LAET 74-289 (Petter 1987, Plate 7.1, fig. 11), partial right mandible with p2–m1 and alveoli for p1 and m2; LAET 75-1959, partial left mandible with p1–p4 and anterior m1 root, teeth in poor condition; LAET 75-2368, partial left mandible with broken m1 and posterior p4 alveolus; LAET 75-2624, partial left mandible with p1–p3, all teeth broken; EP 1169/05, associated left mandible fragments with p2–p3 and broken p4–m2 (partially in matrix). Description: The morphology and dental metrics of this material are in general agreement with living H. ichneumon (Fig. 8.13). The p1 is a small, single-rooted tooth that is approximately equally long and wide. The p2 has two roots and anterior and posterior cingula. Occasionally, there is an indication of a minute posterior accessory cusp in extant individuals. The p3 is two-rooted and has a small posterior cingulum. There are two accessory cusps, a small mesial one a distal one situated between the main cusp and the cingulum. The p4 has two roots. Its cingulum extends distally below the anterior part of the m1 paraconid. Two accessory cusps are present: a somewhat
209
conical-shaped one mesial to the main cusp and a well-developed one distobuccal to the main cusp. The m1 has two roots. The trigonid is well developed and expands buccally at the level of the protoconid. The metaconid is the smallest trigonid cusp and is somewhat lower than the paraconid and protoconid. The talonid is low, less than half the height of the trigonid, and forms a posterior ridge where the individual talonid cusps are indistinguishable. The labial side of the talonid is more robust in comparison with the lingual side. An anterior mental foramen is located beneath p1, and a more posterior one is located under the distal part of the anterior p3 root. Discussion: This material may be separated from specimens of G. pulverulenta by its larger size (Fig. 8.13), a somewhat more robust horizontal ramus, and a less rounded inferior border. The dental morphology and metrics are characteristic of the species H. ichneumon, and the material may be considered morphologically diagnostic. On the basis of present data, it is not possible to distinguish the Pliocene H. ichneumon of Laetoli from the living African H. ichneumon.
Genus Galerella Gray, 1865 The taxonomy and systematics of Herpestinae are not stable, and this is especially true of the position of the nomen Galerella (slender mongooses). A recent study (Veron et al. 2004) came to equivocal conclusions regarding the monophyly of Galerella. We retain Galerella here for practical reasons because it is used by Wozencraft (2005), in what is the most recent and widely used taxonomic compilation of the Carnivora, and because the material we refer to Galerella sp. can be distinguished from material we refer to Herpestes spp. This usage does not imply a specific view of the phylogenetic topology within the Herpestinae. The oldest Galerella thus far described is from Toros Menalla in Chad (ca. 7–6.5 Ma; Peigné et al. 2005). This material, which consists of fragmentary material of three individuals, has been ascribed to the extant species G. sanguinea (Peigné et al. 2005). Today, Galerella includes two species of interest to us: G. sanguinea, which is distributed throughout sub-Saharan Africa (except the Cape Province and surrounding coast in South Africa), and G. pulverulenta, which occupies the southern region where G. sanguinea is absent (Kingdon 1997). Galerella sp.
Fig. 8.13 Bivariate diagram of length and width of m1 in selected Herpestidae. The m1 of LAET 74-289 and that of EP 1169/05 are closely similar to those of H. ichneumon in size and proportions
Specimens: Laetolil Beds, upper unit: LAET 75-2722, partial left mandible with c, p2–p3, and alveoli for p4–m2, c broken; LAET 75-3340 (Petter 1987, Plate 7.2, fig. 10), partial right mandible with p4–m2, p4 broken; LAET 77-4570, partial left mandible with p2–m1 and alveoli for p1 and m2; LAET 78-4691, partial left mandible with p2, broken c, and alveoli for p1 and p3–m1; LAET 78-4955, partial right mandible
210
L. Werdelin and R. Dehghani
with i2–c and p2–p3 (LAET 78-4955a) and partial right mandible with p4 and alveoli for m1 (LAET 78-4955b). Description: The mandibular ramus is slender and straight. There are two mental foramina, the anterior one beneath the p1 alveolus and the posterior one between and ventral to p2 and p3. The p1 was present, as attested to by the presence of its alveolus in some specimens, although this seems variable. The apex of p2 is situated between the two roots. The p2 has two vertically positioned cuspids along the distal edge of the main cusp. The main cusp apex is situated above the anterior half of the tooth. The p3 is somewhat larger than the p2 but has the same basic morphology. The main difference between them is the occurrence of three minute elevations distally along the edge of the main cusp of p3. The distal end of p3 is occupied by a cingulum. The p4 is not well preserved in any specimen. The main cusp is tall, and there is a well developed accessory cusp distobuccal to the main cusp. The trigonid cusps of m1 are well developed and form a nearly equilateral triangle. The distobuccal corner of the paraconid forms a shearing facet together with the mesiobuccal corner of the protoconid. The metaconid is situated distal to the paraconid and slightly lingual to it. The basin-shaped talonid is low and only half the length of the trigonid. It has a posterior, ridgelike development. The proto-, meta-, and entoconid of m2 are all well developed. The protoconid and metaconid form the anterior face of the tooth. The entoconid is located distal to the paraconid. At the posterior end of the talonid there is a ridgelike formation. Discussion: Specimens LAET 75-2722, LAET 75-3340, LAET 77-4570, and LAET 78-4691 resemble the living members of Galerella, and the species G. pulverulenta in particular, in features of the lower dentition. Specimen LAET
78-4955, on the other hand, tends to be somewhat closer to G. sanguinea. The similarity in dental metrics between these specimens from Laetoli and specimens of living Galerella is supported by the morphology of the teeth. Specimen LAET 75-3340, previously considered to belong in the genus Mungos (Petter 1987), tends to be closer to Galerella with regard to dental metrics. The lower carnassial of this specimen is narrower but longer than the lower carnassial of the extant Mungos mungo, as well as specimen LAET 75-3072, a Mungos dietrichi. Another feature specific to M. mungo and M. dietrichi is the increase in mandible depth behind m1 in relation to the length of m1 as compared to other herpestid genera, although there is a slight overlap between M. mungo and G. sanguinea. This increase in the height of mandible is not present in LAET 75-3340. Assignment of LAET 75-2722, LAET 75-3340, and LAET 78-4955 to Galerella is reasonable with regard to dental metrics of the lower jaw (Fig. 8.14) and is supported by the complete reduction of p1. This is also true of LAET 77-4570 and LAET 78-4691, with the exception of the retained first premolar. The presence of p1 is not diagnostic, however, since a first premolar is present in a small number of individuals of living Galerella and may be explained as a primitive retention. There is no positive evidence that all these specimens belong to a single taxon. In fact, the horizontal ramus of specimen LAET 75-2722 is more massive and taller beneath the tooth row in comparison with the other specimens assigned to Galerella sp. With respect to dental metrics, the specimens are intermediate between the smaller Galerella sanguinea and the larger Galerella pulverulenta or are within the size range of the latter species.
Fig. 8.14 Bivariate diagram of length and width of p2 in selected small Carnivora. The p2 of LAET 77-4570, LAET 75-2722, LAET 78-4691, and LAET 79-4955 all match G. sanguinea and H. ichneumon
in proportions, whereas the p2 of M. mungo is broader and that of G. genetta is more slender
8 Carnivora
The identification of Galerella sanguinea from Toros Menalla (Peigné et al. 2005) is based on negative evidence – that is, the absence of characters distinguishing the material from the modern species. We are inclined to doubt this for several reasons, not least because the material discussed here, which is intermediate in age between Toros Menalla and the modern fauna, more extensive, and better preserved, shows characters that do distinguish it from both of the extant species. This suggests that if the Toros Menalla sample were larger, this material would also show distinctive characters.
Genus Helogale Gray, 1862 The genus Helogale (dwarf mongooses) is the most common of the small herpestids in the fossil record, perhaps because of its social habits. Besides Laetoli, the genus is known from Omo, Shungura Fm., Mbs. B, C, E–F, and G, as well as possibly Kanapoi and Hadar, Sidi Hakoma Mb. (Wesselman 1984; Werdelin 2003b). Today, the dwarf mongoose, H. parvula, is distributed from Somalia south along the eastern part of Africa to the Transvaal and west to Namibia and Angola, avoiding the Congo Basin. The Somali dwarf mongoose, H. hirtula, has a distribution from Somalia down through northern Kenya, as far west as the Turkana Basin (Kingdon 1997). Helogale palaeogracilis (Dietrich, 1942) Specimens: Laetolil Beds, lower unit: EP 015/98, left mandible fragment with p4–m1. Laetolil Beds, upper unit: LAET 75-2503 (Petter 1987, Plate 7.2, fig. 13), isolated right P4 with broken mesiobuccal and distal root; LAET 75-2807 (Petter 1987, Plate 7.2, figs. 1 and 2), cranium (LAET 75-2807a) with right and left I1–I3, P2–M2, and mandible (LAET 75-2807b) with right p3–m2, left p2–p4 and m2, right c and some incisors attached to the left mandible fragment; LAET 75-2997, partial left mandible with c, p2–m1 and m2 alveolus; LAET 75-2994 (specimen marked 2994, but labeled 2944), partial right mandible with P3 and broken P4; LAET 75-3368, partial right mandible with c, p2–p3, c and p3 broken, and an additional left mandible fragment with alveolus for m2; LAET 75-3565 (Petter 1987, Plate 7.2, fig. 3), partial left mandible with p3–m1 and m2 alveolus, p3 broken; LAET 75-3616 (Petter 1987, Plate 7.2, fig. 4), partial left mandible with p3–m1, alveoli for p2 and m2; LAET 76-3973, partial left mandible with c, p2–p4 and m1 alveoli, c broken; LAET 78-4736, partial left mandible with c and erupting p3–p4; LAET 78-4980, partial right mandible with p2–p4 and anterior m1 alveolus (given as left fragment in Petter 1987); LAET 78-5295, left mandible fragment with canine root, p3–m1, alveoli for p2 and m2, and an additional left mandible fragment with i2–i3 and partial left c alveolus; LAET 78-5298, partial right mandible (LAET 78-5298a)
211
with p3–p4, anterior portion of m1, p2 alveoli, and left mandible (LAET 78-5298b) with p3–m1, all teeth broken (given as p2–p4 in Petter 1987); EP 436/98, left mandible fragment with c–p3; EP 1787/00, right mandible fragment with c, p2, alveoli for p3, p4–m1, proximal right humerus, occipital fragment, indeterminate postcranial fragments; EP 2577/00, right mandible fragment with root of c, p2–p4; EP 873/03, right mandible fragment with p2–p3; EP 985/03, left mandible fragment with p2–p4; EP 2430/03, right mandible fragment with root of c, p2–m1; EP 993/04, left mandible fragment with p2–m2; EP 1709/04, left mandible fragment with c, alveolus for p1, p2–p3; EP 1097/05, left mandible fragment with c–p2, roots of p3-p4, anterior fragment of m1; EP 1224/05, left mandible fragment with c–p4, anterior half of m1; EP 1790/00, left mandible fragment with c, p3; EP 208/00, right mandible fragment with p3; EP 4167/00, right mandible fragment with roots of p2, p3–p4, roots of m1; EP 348/01, left mandible fragment with p3–m1, left maxilla fragment with C–P2, proximal humerus, proximal ulnae, distal radius, proximal radius, tibia shaft, distal femur fragment, five ribs, six vertebrae, three metapodials, terminal phalanx, cranial and postcranial fragments; EP 467/01, left mandible fragment with p3; EP 390/03, left mandible fragment with p3; EP 770/03, right mandible fragment with p3; EP 041/04, right mandible fragment with alveoli for c, p2, p3–m1; EP 1456/04, left mandible fragment with p3–p4; EP 3858a/00, left mandible fragment with p3, broken p4–m1, complete m2; EP 3858b/00, left mandible fragment with m1–m2 (EP 3858/00 also includes a left edentulous mandible fragment, a right maxilla fragment with P2–P3, proximal humerus, distal humerus, proximal ulna, calcaneum, numerous postcranial fragments); EP 118/05, right mandible fragment with broken p2–p3, complete p4–m1; EP 642/01, right mandible fragment with p4; EP 035/01, right mandible fragment with p4–m2; EP 2888/00, right mandible fragment with p4; EP 2887/00, right mandible fragment with p4–m1; EP 1874/00, right mandible fragment with p4–m1; EP 531/00, right mandible fragment with p4; EP 466/00, left mandible fragment with p4–m2; EP 1500/98, right mandible fragment with p4–m1; EP 4168/00, right mandible fragment with m1; EP 636/01, left mandible fragment with m1. Ndolanya Beds, upper unit: LAET 75-940 (Petter 1987: Plate 2, fig. 12), right maxilla fragment with P3 and P4. Unknown stratigraphic level: LIT 59/359, partial right mandible with p3–m1 and alveolus for m2. Description: The muzzle is short and the braincase relatively elongated. The nasal bones are short, extending ~1.5 mm behind the mid-dorsal part of the frontal-maxillary suture. The sagittal crest consists only of a very short portion merging posteriorly with the nuchal crest. The postorbital processes are well developed but open and short, and they do not extend to the zygomatic arch. The maximum width of the braincase is attained at the posterior part of the zygomatic process. The zygomatic arches are broken on both sides.
212
The mandibular fossa is present on the right side but broken. The tympanic bulla is well developed, with the caudal entotympanic inflated posteroventrally. The entotympanic is C shaped, with the external auditory meatus located at the lateral center of the C arm. There is a minute elevation posteriorly on the canine crown base. There is no alveolus for p1. Between the canine and the p2 there is a small diastema. Both p2 and p3 are oval in occlusal view, with the lingual face somewhat more flattened. The height of these two teeth approximately equals their length. Thickenings of the cingulum can be seen mesiolingually and distally. The distal edge of the cusp is concave, which is more evident in p2 than in p3. In p2, the main cusp apex is located above the mesial root, whereas in p3 the main cusp apex is located between the roots. A distal accessory cusp arises from the distal base of the main cusp of p3. The distal width of p4 is greater than the mesial width. The distal accessory cusp of p4 is well developed, separated from the main cusp, and positioned buccally. A cingulum runs from the mesial edge via the lingual face to the distal edge, where it forms a ridge. The cingulum on p4 forms a small mesiolingual cusplet. In m1, the trigonid is slightly longer than the talonid. The trigonid cusps are well developed. The paraconid slopes down distobuccally in the direction of the protoconid. The protoconid, which is the tallest cusp, is located buccally in the middle of the tooth. The protoconid and metaconid apices are oriented slightly backward. The metaconid is located distal to the paraconid and slightly lingually. The talonid forms a distal ridge, where the individual cusps of the talonid are difficult to distinguish. The distal portion of the trigonid represents the widest part of the tooth at the level of the metaconid and protoconid. There are two mental foramina. The anterior one is located below the mesial margin of the mesial root of p2, whereas the posterior one is located below the mesial root of p3. The ventral border of the horizontal ramus is more or less straight. There is no indication of a P1 alveolus. The upper incisors are small. The buccal surface of the canine is rounded, whereas the lingual surface is flattened. The distal edge is carinate. The P2 is oval in occlusal view, and the cusp is oriented somewhat distally. The apex of the right P2 is located between the two roots, whereas the apex of the left P2 is located closer to the distal root. There are no accessory cusps on P2. The P3 apex is located under the distobuccal root. There is a lingual root on P3 supporting a well-differentiated accessory cusp. The crown of P3 is bordered buccally by a weak cingulum, ending in small cusplets mesially and distally. In both P2 and P3, the height of the tooth approximates the length, and neither tooth has an accessory cusp. All the principal cusps of the upper carnassial are well developed. The protocone is positioned anterolingually. The mesial-most part of the tooth is the base of the parastyle.
L. Werdelin and R. Dehghani
The paracone is situated on a concave diagonal axis between the protocone and metacone. The distal border of the protocone ends at the midpoint of the lingual face of the paracone. The mesiolingual portion of the metacone forms an open angle with the distal part of the paracone. The M1 is reduced, and the width of the tooth is greater than the length. The protocone is located lingually, with the mesiolingual portion of the tooth slightly more mesial than the mesiobuccal portion, here represented by the paracone. The protocone is buccally elongated into two crests, one mesial and one distal, which are almost equal in height to the protocone. The buccal face of the tooth is taller than the lingual face. The M2 is very reduced and small, with no distinguishable cusps. Discussion: Helogale palaeogracilis is the smallest herpestid species from Laetoli. The size of the teeth in the specimens here attributed to H. palaeogracilis is comparable to that of the living dwarf mongooses, H. hirtula and H. parvula (Fig. 8.15). The teeth of extinct members of Helogale are narrower, however, even though they are of approximately the same length as the teeth of their extant relatives. This relative narrowing gives their teeth a more slender appearance. The dental ratios of some teeth of the living dwarf mongooses, H. hirtula and H. parvula, differ from the ratios seen in H. palaeogracilis. The length of p3 in H. hirtula does not differ from specimens described as H. palaeogracilis (Petter 1987). However, p3 in H. hirtula is considerably broader in comparison with the Laetoli specimens. This pattern is also seen in the lower carnassial. The lower carnassial of H. parvula
Fig. 8.15 Bivariate diagram of length and width of m1 in selected Herpestidae. Despite the poor sample of extant Helogale, it can be seen that the fossil specimens referred to this genus are all in the size range of the extant species. The considerable range of variation in width of the fossil sample may be due to the presence of more than one species-level taxon or (more likely) to measurement error resulting from variable preservation of the material
8 Carnivora
shows the opposite pattern to that seen in H. hirtula. There is a length reduction in the m1 of H. parvula in comparison with the material from Laetoli, whereas there is no difference in mean absolute tooth width. The Laetoli specimen LAET 75-940, referred to Cynictis by Petter (1987), has relatively small upper premolars, similar to those of LAET 75-2807 and LAET 75-2994, and is clearly distinct from the yellow mongoose, Cynictis penicillata. These bivariate diagrams illustrate the overall smaller size of the P3 in LAET 75-940, LAET 75-2807, and LAET 75-2994 and of the P4 in LAET 75-2503, LAET 75-2867, and LAET 75-940 in comparison with extant Cynictis material. The small posterior elevation on the lower canine of specimen LAET 76-3973 is present in specimen NRM VE A583001, an extant H. parvula, but is not seen in Cynictis. Laetoli specimen LAET 75-2991 has a relatively small p4, similar to specimens attributed to H. palaeogracilis (Petter 1987). The mandible of LAET 75-2991 is also relatively low behind m1, which sets it apart from larger genera, such as Galerella and Herpestes. The majority of specimens referred to the taxon H. palaeogracilis in the present study were already included in this taxon by Petter (1987). The fossil sample of Helogale is indistinguishable from modern Helogale spp. in many morphological respects but differs in generally having narrower cheek teeth. Petter (1987) considers H. palaeogracilis to be a Helogale species with primitive dental characters reminiscent of Galerella, although apomorphic cranial characters confirm its assignment to Helogale.
Helogale cf. H. palaeogracilis Specimens: Laetolil Beds, upper unit: LAET 75-399 (Petter 1987, Plate 7.2, fig. 9), partial right mandible with c and broken m1, alveoli for p2–p4 and m2; LAET 75-405, partial right mandible with broken p3–p4, alveoli for p2 and m1–m2; LAET 75-1974, partial left mandible with c and broken p2–p3; LAET 75-3334, left mandible fragment with anteriorly damaged p4 and roots of c, p2–p3 and alveoli for m1–m2; EP 1324/04, right mandible fragment with p4–m1. Description: Specimen LAET 75-1974 is metrically similar to the dwarf mongoose H. parvula. The canine and the third premolar in the lower jaw are narrower and shorter in H. parvula, LAET 75-1974, LAET 75-2991, and LAET 76-3973 than in the remaining specimens. Discussion: The complete reduction of p1, the minute posterior elevation (weak in specimens LAET 75-399 and LAET 75-1974 because of preservation) on the canine crown base, and the size of the mandible indicate that this material may be referred to H. palaeogracilis, but this attribution must remain tentative given the condition of the material.
213
Genus Mungos Geoffroy St.-Hilaire and Cuvier, 1795 Specimens referred to the genus Mungos (banded mongooses) are known from a few eastern African localities apart from Laetoli. The species M. dietrichi is tentatively known from the middle and upper parts of the Lomekwi Mb. of the Nachukui Fm., West Turkana, and the extant species M. mungo is known from Olduvai, Bed I (Petter 1973, 1987). Today, the genus comprises two species: the banded mongoose, M. mungo, which is distributed throughout eastern and central Africa, except in densely forested regions, and the southern margin of the Sahara in western Africa; and the Gambian mongoose, M. gambianus, which is known from savannas and woodlands in western Africa from Senegal to the Niger River (Kingdon 1997).
Mungos dietrichi Petter, 1963 (Fig. 8.16) Specimens: Laetolil Beds, upper unit: LAET 75-2769, partial right mandible with p3–p4, posterior p2 and anterior m1 alveoli (given as left in Petter 1987); LAET 75-3741 (Petter 1987, Plate 7.2, fig. 11), partial left maxilla with P4–M1; LAET 75–548, partial left mandible with broken m1 and posterior p4 alveolus; LAET 77–4571 (Fig. 8.16a–c), partial right mandible with p4–m2. Ndolanya Beds, upper unit: LAET 75-3072 (Petter 1987, Plate 7.1, fig. 2, Plate 7.2, fig. 14), left mandible with c, p2–m2 and alveolus for p1; EP 1217/03, right mandible fragment with p4–m1, alveolus for m2. Description: The ventral border of the mandible is convex. The angular process is long. The anterior border of the coronoid process slopes posteroventrally immediately behind the second lower molar. High crowns and sharp cusps characterize the teeth, particularly the premolars. The canine is tall, and the apex points dorsally. The lingual surface of the canine is flattened and the buccal surface convex. There is a small alveolus for p1. The p2 has two roots, and the apex is above the mesial half of the tooth. The basal contour of p2 is oval and slightly broadened posteriorly, with an incipient cingulum formation at the posterolingual angle. There are no accessory cusps on p2. The p3 also has two roots. The apex of the main cusp of p3 is located above the juncture between the two roots. The main cusp is conical, mesially placed, and buccally flanked by a distal accessory cusp. A distal cingulum runs from the distal part of the accessory cusp to the distolingual angle of the tooth. Like p2 and p3, p4 has two roots. The basal contour of this tooth is rectangular, with rounded corners. The main cusp is pointed and conical. There is a well-developed accessory cusp distobuccally and a cingulum distally. The lower carnassial
214
Fig. 8.16 (a–c) Mungos dietrichi LAET 77-4571; right mandibular ramus in (a) buccal, (b) lingual, and (c) occlusal view. (d–f) Mungos sp. nov.?, LAET 75-1923; right mandibular ramus in (d) buccal, (e) lingual, and (f) occlusal view
has two roots and a well-developed trigonid. The trigonid is marginally longer than the talonid. All trigonid cusps are more or less conical in shape and are located close to each other, especially the paraconid and metaconid. The paraconid is located at the mesiolingual corner in front of the metaconid. The protoconid is situated somewhat behind the paraconid and slightly in front of the metaconid on the buccal side. The paraconid and metaconid are joined at their bases, whereas the protoconid is distinct. The talonid of m1 is shorter and lower than the trigonid and square in outline. The hypoconid is large and well developed but worn. The distal border of the talonid forms a ridge surrounding the talonid depression immediately posterior to the trigonid. The m2 somewhat overlaps the distolingual corner of the m1 talonid. All m2 cusps are distinguishable despite being reduced. The protoand metaconid of m2 are relatively well developed, whereas the paraconid is present as a small bump between the two other cusps. The talonid of m2 is somewhat smaller than the trigonid. The hypoconid is the most readily distinguishable
L. Werdelin and R. Dehghani
cusp of the m2 talonid, whereas the hypoconulid and entoconid are merged into a posterior crest. The upper carnassial has three roots and a triangular basal contour with a salient protocone. Unfortunately, the distal part of the P4 metastyle is broken in the only available specimen. In this upper carnassial, the width of the tooth exceeds its length. The salient protocone is well developed, conical, and wide at its base. The parastyle is wide and conical, similar to the protocone but much smaller. The paracone is slightly larger than the protocone. The metastyle blade is somewhat reduced and more buccally oriented in comparison with the parastyle and paracone. The M1 has three roots; the two buccal roots are probably small. The mesiobuccal corner where the paracone is located is damaged. The M1 is wider than it is long. The protocone is well developed and salient. The metacone is somewhat reduced. Discussion: The teeth of the extant banded mongoose, Mungos mungo and specimens herein referred to M. dietrichi, are wider relative to length than those of Helogale, Galerella, and Herpestes. The length of the teeth in LAET 75-3072, the best preserved specimen of M. dietrichi from Laetoli, is similar to that of a large Galerella, whereas the width of the teeth is more like that of the smaller Herpestes, making the teeth in LAET 75-3072 relatively wide in relation to their length, as also seen in M. mungo. However, the ratios of tooth lengths in LAET 75-3072 and M. mungo do not deviate from the pattern seen in the rest of the comparative sample. A feature diagnostic of Mungos is the salient protocone on the upper carnassial, sometimes resulting in the width of the carnassial exceeding its length. This feature is seen in extinct as well as extant specimens of the genus. This increase in carnassial tooth width is particularly evident in specimen LAET 75-3741 from Laetoli. The most complete dental series of M. dietrichi comes from mandible LAET 75-3072. This specimen approaches and sometimes even exceeds M. mungo in size; this is also true of specimen LAET 75-2769. A difference between M. mungo and M. dietrichi is the presence of p1 in the latter species, as seen in mandible LAET 75-3072 from Laetoli, as well as mandible FLK N 6128 from Olduvai (Petter 1973).
Mungos sp. nov.? (Fig. 8.16) Specimens: Laetolil Beds, upper unit: LAET 75-1923 (Fig. 8.16d–f), partial right mandible with p4–m1 and alveolus for m2; EP 544/01, left mandible fragment with m1–m2. Description: The anterior portion of p4 is broken. Distobuccal to the main cusp there is an accessory cusp, of which the uppermost part of the apex is broken. There is a posterior cingulum. The widest part of p4 is probably at the level of the accessory cusp, since it protrudes from the buccal face of the tooth. The paraconid is slightly the taller of the trigonid cusps on m1, but the protoconid is by far the largest. The mesial and
8 Carnivora
particularly the mesiolingual portions of the trigonid are strongly compressed, making the distance between the paraconid and metaconid apices very short, approximately equal to the total length of the protoconid. The talonid is slightly shorter and much lower than the trigonid. The entoconid is well developed, whereas the hypoconid and hypoconulid are small, low, and set close together. The m2 is a large, well-developed tooth. The protoconid and metaconid are both well developed and set widely apart. The entoconid is large, as is the hypoconid, which is separated from the metaconid by a deep postvallid notch. The hypoconulid is set posteriorly and is ridgelike. Discussion: These specimens resemble M. mungo and M. dietrichi in general features but differ from them in the morphology of m1 and the mandible posterior to the tooth row. They are likely to belong in Mungos but may represent a new species within that genus. A larger sample of fossil Mungos spp. is required to address this question.
Family Hyaenidae Gray, 1821 The fossil record of Hyaenidae in Africa is extensive (summarized in Werdelin and Turner 1996). All the living species evolved on this continent, which, therefore, is key to understanding the evolution of the extant Hyaenidae. The diversity within the family was greater in the past, and Laetoli exemplifies this, with six species in as many genera. Metric data for Laetoli Hyaenidae are given in Tables 8.5 and 8.6.
Genus Crocuta Kaup, 1828 Crocuta is common in the late Pliocene and Pleistocene of Africa. A number of taxa are involved, and the evolution of the genus seems linked to that of the hominin lineage (Lewis and Werdelin 2000). The record of C. dietrichi from the Laetolil Beds, upper unit, is the oldest material of the genus, though material from the Kataboi Fm., West Turkana, which belongs to a distinct, undescribed taxon, may be of approximately the same age. Crocuta dietrichi Petter and Howell, 1989 Specimens: Laetolil Beds, upper unit: LAET 75-2953 (holotype, Barry 1987, fig. 7.9a and b; Petter and Howell 1989, fig. 1c and d), left mandible fragment with p2–m1; LAET 76-3970/77-5370, right mandible fragment with p2–p4, isolated left p2, p4; EP 1067/04, left mandible fragment with c root, broken p2–p3, p4 roots; LAET 74-185 (Barry 1987, fig. 7.9c [as LAET 158]; Petter and Howell 1989, fig. 1a and b), left maxilla fragment with P1-P4; LAET 74-149, right
215
mandible fragment with p2–p3. Ndolanya Beds, upper unit: LAET 78-5107 (Barry 1987, fig. 7.9d), right mandible fragment with p2–p4, anterior half of m1; LAET 76-3951 (Barry 1987, fig. 7.11), left mandible fragment with p2–p4; EP 1390/05, right mandible fragment with m1. Description: Note that although the description of the mandible and lower dentition is a composite, that of the maxilla and upper dentition is based on the single specimen LAET 74-185. The mandibular ramus is robust. The height of the mandibular ramus increases from mesial (moderately tall) to distal (very tall). There is a single mental foramen located beneath the middle of p2. The masseteric fossa extends mesially to about the distal end of m1. The symphysis is large. The rugosity ends just mesial to p2, but the flattened area continues to beneath the middle of p3. The lower canine is known only from the root, which is a rounded oval in cross-section, with its longest diameter at approximately 30° to the cheek tooth row. It is followed by a diastema of a little less than 10 mm. The p2 is robust but low. There is no mesial accessory cusp; the main cusp is low, and the distal accessory cusp is low, somewhat trenchant, and set directly behind the main cusp. The p3 is pyramidal. The tooth is very broad mesially, but there is no mesial accessory cusp. The crest on the mesial face of the main cusp is prominent. The whole tooth has a distalward slant. The distal accessory cusp is small, short, and appressed to the main cusp. It is flanked on either side by narrow shelves. The p4 is robust and wide. The mesial accessory cusp is small and appressed to the main cusp. The main cusp is pyramidal and short. The distal shelf is long but relatively narrow, and its lingual side is formed into a low crest. The m1 is long and low. The paraconid is longer than the protoconid. The metaconid is either very small or simply a bump on the distolingual side of the protoconid. The talonid is short. It has one or two very small cusps. The P1 is a small, single-rooted tooth. The P2 is short and robust. It has a pyramidal crown and a distal (but no mesial) accessory cusp. The P3 is short and wide, with a pyramidal crown. The mesiolingual accessory cusp is small. The crest from it to the apex of the main cusp is strongly developed. The distal accessory cusp is low and appressed to the main cusp. The P4 is long and slender. The parastyle is strong. The protocone is well developed but low. The paracone is broken but must have been tall, whereas the metastyle is long and low, turning buccally at its distal end. Discussion: C. dietrichi was first described as “Crocuta new species?” by Barry (1987), then formally described by Petter and Howell (1989). The main characteristic differentiating this species from modern C. crocuta, according to these authors, is the reduced size, and especially the small premolars. Barry (1987) lists a number of other characters, some of which are difficult to assess because of the small sample size, some of which fall within the range of variation of the modern species, and yet others whose diagnostic value has been enhanced by the discovery of additional specimens of
216
L. Werdelin and R. Dehghani
Fig. 8.17 Bivariate diagram of length and width of p4 in selected Hyaenidae. Note the slender p4 of C. crocuta compared with those of other Crocuta. C. dietrichi from Ahl al Oughlam was described as C. dbaa by Geraads (1997)
C. dietrichi from Laetoli and other East African sites in the age range of 3.5–2.5 Ma, such as Koobi Fora, West Turkana, and Omo. The youngest record of C. dietrichi is from the Upper Burgi Mb. at Koobi Fora – that is, latest Pliocene. Many of these characters are useful in distinguishing C. dietrichi from other fossil African Crocuta. However, the most significant character distinguishing C. dietrichi from modern Crocuta has not been mentioned by previous authors. This is the relatively broad p4 (Fig. 8.17), a feature shared with other fossil African Crocuta, mainly C. ultra, whereas in the modern species, the p4 is quite slender. This, together with the relatively long talonid, provides a stable characterization of the species C. dietrichi, contra the opinion of Turner (1990), who considered the latter to fall within the range of variation of the modern species.
Genus Parahyaena (Hendey, 1974) Parahyaena was named as a subgenus of Hyaena by Hendey (1974) in recognition of differences between the brown and striped hyenas. Although it was originally restricted to the extant brown hyena, Werdelin (2003b) included the fossil P. howelli from Kanapoi in the genus. The Laetoli material represents the second known fossil record of this genus.
Parahyaena howelli Werdelin, 2003 (Fig. 8.18) Specimens: Laetolil Beds, lower unit: KK 82-58 (Fig. 8.18), partial skeleton including cranium, mandibles and nearly
complete dentition. Laetolil Beds, upper unit: EP 463/01, left mandible fragment with c, p2-p3 - p4-m1 lost and alveoli resorbed; EP 395/98, isolated right p3; EP 829/00, isolated left c, p2, p3; LAET 76-4008a (Barry 1987, fig. 7.12), right maxilla fragment with P4. Tentatively assigned specimens: Laetolil Beds, upper unit: LAET 76-4092, isolated right P3; LAET 76-4008b, isolated left P2. Description: The mandibular ramus is robust and deep, with a distinct ventral angle beneath m1. There is a single mental foramen located beneath p2. The coronoid process is tall, and its dorsal end is squared off. The masseteric fossa is deep and reaches mesially to approximately the m2. The i2 is a small, spatulate tooth without any distinct lingual accessory cusp. The i3 is somewhat larger but still incisiform. It is separated from the lower canine by a diastema of approximately 3 mm. The lower canine is set at approximately 30° to the main axis of the tooth row. It is a flattened oval lacking distinct mesial and distal keels. The postcanine diastema is approximately 10 mm long. The p2 is short. There is no mesial accessory cusp, but a small distal one is set in a short distal shelf. The main cusp is low. The p3 is robust with a squared-off mesial end lacking an accessory cusp. The main cusp is tall and pyramidal, whereas the distal accessory cusp is low and set free of the main cusp. The p4 is long and slender. It has a very low mesial accessory cusp, a low main cusp, and a low, long distal accessory cusp that is set free of the main cusp. The distolingual ridge is short. The m1 is relatively short. The paraconid is somewhat longer than the protoconid, and there is a small metaconid. The talonid is short and has two small cusps that are probably the hypoconid and entoconid. The m2 is a small, single-rooted tooth without
Fig. 8.18 Parahyaena howelli, KK 82-58. Skull in (a) right lateral and (b) ventral view. Left mandibular horizontal ramus in (c) buccal, (d) occlusal, and (e) lingual view
218
distinct cusps. The upper incisors are set along an arc. The I1 and I2 are small, spatulate teeth, with the I2 very slightly the larger of the two. The I3 is not preserved, but the alveolus indicates that it was about twice the size of I1 and I2. To judge from their alveoli, the upper canines were flattened ovals in occlusal view. There is no P1. The P2 is small and slender. It has no mesial accessory cusp, a low main cusp, and a well-developed distal accessory cusp. The P3 is rounded in occlusal view. Its mesial end is slender, and there is a small accessory cusp set mesial and slightly lingual to the main cusp. The main cusp is pyramidal. There is a distinct cingulum shelf on its lingual side, so that the tooth is widest there. The P4 is long and robust. The parastyle is well developed and slightly larger than the protocone. The latter is set slightly in front of or in line with the parastyle. The paracone is tall and mesiodistally short. The metastyle is longer than the paracone and straight, bending buccally only in its last few millimeters. The M1 is preserved only as a fragment. It was mesiodistally short and buccolingually long. Discussion: The specimens listed above differ from those assigned to C. dietrichi in a number of ways. Most notably, the premolars are considerably more slender, with the p4 approaching C. crocuta in this feature (Fig. 8.17). On the other hand, they are closely similar to the hypodigm of Parahyaena howelli from Kanapoi (Fig. 8.17; Werdelin 2003b) and they can safely be assigned to this taxon, possibly extending the known temporal range of this species by as much as half a million years.
Genus Ikelohyaena Werdelin and Solounias, 1991 Werdelin and Solounias (1991) named Ikelohyaena for the Hyaena abronia of Hendey (1974), recognizing its distinction from the striped hyenas of the genus Hyaena. Ikelohyaena is known from a number of sites in Africa, of which Ahl al Oughlam, where “?Hyaenictitherium barbarum” was found (Geraads 1997), is the youngest.
Ikelohyaena cf. I. abronia Hendey, 1974 Specimens: Laetolil Beds, upper unit: LAET 75-3338 (Barry 1987, fig. 7.10), left mandible fragment with p2–m1; LAET 75-1849, isolated left p4; EP 1046/98, left maxilla fragment with P2–P4. Unknown stratigraphic level: LIT 59/465, left maxilla fragment with P3, anterior part of P4. Tentatively assigned specimens: Ndolanya Beds, upper unit: EP 1218/03, left maxilla fragment with I3 alveolus, C, P1 root, P2.
L. Werdelin and R. Dehghani
Description: The mandibular corpus is robust, with a fairly straight ventral border. There is a single mental foramen situated beneath the middle of p2. The masseteric fossa extends mesially to approximately the level of m2. The p2 has a more or less rectangular occlusal outline. It has no mesial accessory cusp and only a small distal one. The main cusp is robust but relatively low. The p3 is similar to p2 but is much larger. Its main cusp is located toward the mesial end of the tooth, and the mesial accessory cusp is very small. The distal part of the tooth is broken, and the shape and structure of the distal accessory cusp and shelf cannot be determined. Overall, the tooth greatly resembles the p3 of Hyaena. The p4 is a relatively short, slender tooth. The mesial accessory cusp is very small, whereas the main cusp is tall and pyramidal. The distal accessory cusp is prominent and set free of the main cusp, not appressed to it. The distolingual shelf has a blunt crest and broadens out on the lingual side. The infraorbital foramen of the maxilla is located above P3. The P2 is similar in structure to p2 but has a prominent cingulum on its lingual side. There are no accessory cusps, and the main cusp is relatively low. The P3 is a robust tooth with a very small mesiolingual accessory cusp; a tall, pyramidal main cusp; and a prominent but low distal accessory cusp appressed to the main cusp. There is a strong lingual cingulum at the base of the tooth. The P4 is short and robust. The parastyle and protocone are both strongly developed and set approximately level with each other. The paracone is tall, whereas the metastyle is broken but was apparently low. Discussion: This material strongly resembles the topotypic material of I. abronia from Langebaanweg (Hendey 1978; Werdelin et al. 1994). This was also noted by Barry (1987) in his comparison between LAET 75-3338 and Hyaenictis preforfex (= I. abronia). This species has a pivotal role in the evolution of modern hyaenas. Its extensive stratigraphic range, from the late Miocene to the late Pliocene, as well as its intermediate morphological features, indicate that it is either the first species to evolve (on the Hyaena lineage) after the split between the striped and brown hyenas (genera Hyaena and Parahyaena, respectively) or is the last common ancestor of these two. Resolving this issue will greatly assist in understanding the factors leading to the evolution of the modern scavenging hyenas. The Upper Ndolanya Beds material of Ikelohyaena is among the last of the species, which has its last known occurrence at Ahl al Oughlam in Morocco (?Hyaenictitherium barbarum in Geraads 1997). Laetoli is thus far the only site where Ikelohyaena co-occurs with Parahyaena howelli, suggesting that the niches of these fossil precursors may have been more different than the niches of their modern descendants, the striped and brown hyenas, which do not have overlapping ranges.
8 Carnivora
Genus Lycyaenops Kretzoi, 1938 The genus Lycyaenops was created by Kretzoi (1938) for L. rhomboideae from the Pliocene of Hungary. Because of the sketchy nature of Kretzoi’s description of and figure illustrating this taxon, doubt remained concerning its validity and phylogenetic position until Werdelin (1999a) showed that it was a member of the “hunting hyena” lineage that also includes Lycyaena, Hyaenictis, and Chasmaporthetes (Werdelin et al. 1994). Lycyaenops cf. L. silberbergi (Broom in Broom and Schepers, 1946) (Fig. 8.19) Specimens: Laetolil Beds, upper unit: LAET 75-494 (Fig. 8.19), right maxilla fragment with P2–P3. Unknown stratigraphic level: NHM AS 7.VI.35, left P3. Description: The P2 is slender and rectangular, with its distal end only slightly broader than the mesial end, and the middle of the tooth has a shallow but distinct “waist.” There is no distinct mesial accessory cusp. The main cusp is tall and trenchant. The distal accessory cusp is large, somewhat trenchant, and free of the main cusp. It is bordered lingually by a narrow shelf. The P3 has a small mesial accessory cusp set at the mesiolingual corner. The main cusp was tall. There is a basal cingulum distal to the mesial accessory cusp on the lingual side of the tooth. In the isolated P3, the main cusp is very tall and trenchant, and the distal accessory cusp is prominent (Turner 1990, fig. 2).
219
Discussion: Turner (1990) recorded the presence of Chasmaporthetes silberbergi from an unknown level at Laetoli, based on a P3 in the collections of the Natural History Museum, London. Subsequently, Werdelin (1999a) transferred this species to the genus Lycyaenops, also in the Lycyaena lineage of “hunting hyaenas.” Specimen LAET 75-494 shows a number of characters indicating that it belongs in this lineage, particularly the relatively slender and long P2 (Fig. 8.19), as noted by Barry (1987). The latter author expressed reservations about assigning the specimen to the Lycyaena lineage on the basis of the weak mesial accessory cusps of the P2 and P3. However, these cusps are much less well developed in Lycyaenops than in either Lycyaena or Chasmaporthetes. Therefore, we conclude that this specimen likely represents a species of Lycyaenops. Neither this specimen nor the one in the Natural History Museum, London, is adequate to definitively record the presence of the species L. silberbergi at Laetoli, however.
Genus ?Pachycrocuta Kretzoi, 1938 Pachycrocuta, which is a mainly Eurasian genus (Turner and Antón 1996), has long been known from several finds in South Africa but was not definitely recorded from eastern Africa until material from West Turkana was described by Werdelin (1999b). Some of this material is only slightly younger than the upper unit of the Laetolil Beds. ?Pachycrocuta sp. (Fig. 8.20) Specimens: Laetolil Beds, upper unit: EP 1370/00 (Fig. 8.20), fragment of p3 or P3. Description: This specimen is a small fragment of the ?anterobuccal part of a third premolar including an enamel fragment and partial root. The shape of the cingulum, the rounding of the tooth and the texture of the enamel all indicate that it belongs in the Hyaenidae. Discussion: The specimen is much larger than the third premolars of any of the other hyaenid taxa from Laetoli and only matches Pachycrocuta in size. Therefore, we tentatively record the presence of this genus in the Laetolil Beds. This taxon is known from roughly contemporaneous sites in West Turkana (Werdelin 1999b).
Genus Proteles Geoffroy St.-Hilaire, 1824 Fig. 8.19 Lycyaena silberbergi, LAET 75-494, maxilla fragment in (a) buccal, (b) lingual, and (c) occlusal view
The fossil record of aardwolves has been limited to a few records from the Pleistocene of South Africa (Werdelin and
220
Fig. 8.20 ?Pachycrocuta sp., EP 1370/00, dental fragment in buccal? view
L. Werdelin and R. Dehghani
distal right humerus fragment; KK 82-35, distal right radius fragment; KK 82-36, ulna distal shaft fragment; KK 82-37, ulna right shaft fragment; KK 82-38, radius? shaft fragment; KK 82-54, distal humerus shaft fragment. These elements likely belong to a single individual Description: The proximal humerus is small, approximately 2.5 cm wide transversely and 3 cm long anteroposteriorly. The greater tubercle rises considerably above the head, whereas the lesser tubercle is weak and does not quite rise to the level of the head. The head is approximately equally long and wide. The distal humerus is narrow, indicating cursorial adaptations. The medial and lateral epicondyles are well developed but not prominent. Discussion: This material is limited and poorly preserved, but the humerus fragments are clearly identifiable as hyaenid on the basis of the size and shape of the greater trochanter and the shape of the medial and lateral distal condyles. All the material matches modern Proteles in size. Unfortunately, no craniodental material has been preserved. Though no specific diagnostic features of the humerus of Proteles are known, there is no other post-Miocene hyaenid of this size. Hence, we here suggest that the material represents a hyaenid in the Proteles lineage, if not Proteles itself. If confirmed, this would represent the oldest Proteles known, though the lineage itself is far older, at least 10–11 million years (Koepfli et al. 2006).
Family Felidae Fischer, 1817 The Felidae has an extensive fossil record in Africa. However, this record is strongly biased toward the sabertooth forms (Machairodontinae), especially the genera Dinofelis and Homotherium (Werdelin and Lewis 2001, 2005). Conicaltoothed cats are generally much less well represented. However, Laetoli is an exception to this rule, with conical-toothed cats represented by extensive material of five or six species and the sabertooths by limited material of just two. Metric data for Laetoli Felidae are given in Tables 8.7 and 8.8.
Fig. 8.21 aff. Proteles sp. (a) KK 82-32, proximal right humerus in lateral view. (b) KK 82-33, distal right humerus in anterior view
Solounias 1991). The extinct species P. amplidentus from the late Pliocene of Swartkrans differs from the extant species in having slightly less reduced cheek teeth, but the specific distinction is debatable. aff. Proteles sp. (Fig. 8.21) Specimens: Laetolil Beds, lower unit: KK 82-32 (Fig. 8.21a), proximal right humerus fragment; KK 82-33 (Fig. 8.21b),
Genus Dinofelis Zdansky, 1924 The genus Dinofelis is the most common cat genus in the Plio-Pleistocene of Africa. It was revised by Werdelin and Lewis (2001). Dinofelis petteri Werdelin and Lewis, 2001 Specimens: Laetolil Beds, upper unit: LAET 75-448, left P4; LAET 78-4812, left distal radius fragment; LAET 75-868, right i3; LAET 78-5015, left p4 fragment. Ndolanya Beds, upper unit: LAET 78-5045, left distal humerus fragment.
8 Carnivora
Description: This material was described by Werdelin and Lewis (2001), and no further material has been recovered since. These descriptions therefore need not be reiterated here. Discussion: A discussion of the status and affinities of this material was provided by Werdelin and Lewis (2001). Dinofelis petteri is known from a number of other early Pliocene East African localities, such as Kanapoi, Allia Bay, and Hadar.
Genus Homotherium Fabrini, 1890 Homotherium is present in small numbers at most PlioPleistocene African sites. Homotherium sp. (Fig. 8.22) Specimens: Laetolil Beds, upper unit: LAET 75-2028, anterior fragment of right P4; LAET 74-259, left P4 metastyle fragment; LAET 75-2371, left P4 metastyle; LAET 78-4658, left i3; LAET 75-992, left p3; LAET 78-4977, associated left i1, i2, and i3; EP 1044/98, left proximal MT II; EP 1227/98, proximal MT III; EP 575/00 (Fig. 8.22), left and right C; EP 2545, left astragalus. Ndolanya Beds, upper unit: EP 2197/00, left dc; EP 1216/03, right i3.
221
Description: The incisors all are of typical Homotherium morphology, sharply pointed with small medial and larger lateral accessory cusps. The p3 is a small single-rooted tooth with no mesial accessory cusp and only a tiny distal one. The main cusp is low. The upper canines are long, laterally flattened, and strongly curved (Fig. 8.22). The mesial and distal serrations typical of Homotherium have been obliterated by wear. The P4 has a prominent parastyle and incipient preparastyle. The protocone is nearly completely reduced. The paracone is relatively low. The metastyle is low and long with a curve in the middle as is typical of Homotherium. Discussion: Homotherium can be recognized on the basis of numerous characters of the skull, dentition, and postcranium. All of the characters listed above show features that definitely ally them with this genus. On the other hand, the species-level taxonomy of Homotherium in Africa has not been explored in any detail, although several species have been named, such as H. problematicum from Makapansgat (Collings 1972) and H. hadarensis from Hadar (Petter and Howell 1988). Therefore, it is at present not possible to identify the Laetoli Homotherium to the species level.
Genus Panthera Oken, 1816 The genus Panthera represents a conundrum in the evolution of Felidae. According to molecular studies (Johnson et al. 2006), the genus has a divergence time of >10 Ma, yet the earliest appearance of Panthera is from the upper unit of the Laetolil Beds at